List of AMD accelerated processing units

This article uses HTML markup. Please help by changing HTML markup to Wiki markup where appropriate. For help finding or replacing the problematic tags, see instructions. (February 2019)

This is a list of microprocessors designed by Advanced Micro Devices, under the AMD Accelerated Processing Unit product series.

Features overview[edit]

The following table shows features of AMD's APUs

Codename	Server	Basic						Toronto
	Server	Micro														Kyoto
	Desktop	Performance										Renoir	Cezanne
		Mainstream	Llano	Trinity	Richland	Kaveri	Kaveri Refresh (Godavari)	Carrizo	Bristol Ridge	Raven Ridge	Picasso	Renoir	Cezanne
		Entry	Llano	Trinity	Richland	Kaveri	Kaveri Refresh (Godavari)	Carrizo	Bristol Ridge	Raven Ridge	Picasso
		Basic														Kabini
	Mobile	Performance										Renoir	Cezanne
		Mainstream	Llano	Trinity	Richland	Kaveri		Carrizo	Bristol Ridge	Raven Ridge	Picasso	Renoir	Cezanne
		Entry	Llano	Trinity	Richland	Kaveri		Carrizo	Bristol Ridge	Raven Ridge										Dalí
		Basic													Desna, Ontario, Zacate	Kabini, Temash	Beema, Mullins	Carrizo-L	Stoney Ridge	Dalí
	Embedded			Trinity		Bald Eagle		Merlin Falcon, Brown Falcon		Great Horned Owl		Grey Hawk			Ontario, Zacate	Kabini	Steppe Eagle, Crowned Eagle, LX-Family		Prairie Falcon	Banded Kestrel
Platform			High, standard and low power												Low and ultra-low power
Released			Aug 2011	Oct 2012	Jun 2013	Jan 2014	2015	Jun 2015	Jun 2016	Oct 2017	Jan 2019	Mar 2020	Jan 2021 (mobile); Apr 2021 (Desktop)[1]		Jan 2011	May 2013	Apr 2014	May 2015	Feb 2016	Apr 2019
CPU microarchitecture			K10	Piledriver		Steamroller		Excavator	"Excavator+"^[1]	Zen	Zen+	Zen 2	Zen 3		Bobcat	Jaguar	Puma	Puma+^[2]	"Excavator+"	Zen
ISA			x86-64												x86-64
Socket	Desktop	High-end	N/A												N/A
		Mainstream	N/A					AM4
		Entry	FM1	FM2		FM2+^[a]			N/A
		Basic	N/A												N/A	AM1	N/A
	Other		FS1	FS1+, FP2		FP3		FP4		FP5		FP6			FT1	FT3	FT3b		FP4	FP5
PCI Express version			2.0			3.0									2.0				3.0
Fab. (nm)			GF 32SHP (HKMG SOI)			GF 28SHP (HKMG bulk)				GF 14LPP (FinFET bulk)	GF 12LP (FinFET bulk)	TSMC N7 (FinFET bulk)			TSMC N40 (bulk)	TSMC N28 (HKMG bulk)	GF 28SHP (HKMG bulk)			GF 14LPP (FinFET bulk)
Die area (mm²)			228	246		245		245	250	210^[3]		156	156		75 (+ 28 FCH)	107		?	125	149
Min TDP (W)			35	17				12				10			4.5	4	3.95	10	6
Max APU TDP (W)			100			95		65							18	25
Max stock APU base clock (GHz)			3	3.8	4.1	4.1		3.7	3.8	3.6	3.7	3.8	3.9		1.75	2.2	2	2.2	3.2	3.3
Max APUs per node^[b]			1												1
Max CPU^[c] cores per APU			4									8			2	4			2
Max threads per CPU core			1							2					1					2
Integer structure			3+3	2+2						4+2		4+2+1	4+2+1		1+1+1+1				2+2	4+2
i386, i486, i586, CMOV, NOPL, i686, PAE, NX bit, CMPXCHG16B, AMD-V, RVI, ABM, and 64-bit LAHF/SAHF
IOMMU^[d]			N/A
BMI1, AES-NI, CLMUL, and F16C			N/A												N/A
MOVBE			N/A												N/A
AVIC, BMI2 and RDRAND			N/A												N/A
ADX, SHA, RDSEED, SMAP, SMEP, XSAVEC, XSAVES, XRSTORS, CLFLUSHOPT, and CLZERO			N/A												N/A
WBNOINVD, CLWB, RDPID, RDPRU, and MCOMMIT			N/A												N/A
FPUs per core			1	0.5						1					1				0.5	1
Pipes per FPU			2												2
FPU pipe width			128-bit									256-bit			80-bit	128-bit
CPU instruction set SIMD level			SSE4a^[e]	AVX				AVX2							SSSE3	AVX			AVX2
3DNow!			3DNow!+	N/A											N/A
PREFETCH/PREFETCHW
FMA4, LWP, TBM, and XOP			N/A							N/A					N/A					N/A
FMA3			N/A												N/A
L1 data cache per core (KiB)			64	16				32							32
L1 data cache associativity (ways)			2	4				8							8
L1 instruction caches per core			1	0.5						1					1				0.5	1
Max APU total L1 instruction cache (KiB)			256	128		192				256		512			64	128			96	128
L1 instruction cache associativity (ways)			2			3				4		8			2				3	4
L2 caches per core			1	0.5						1					1				0.5	1
Max APU total L2 cache (MiB)			4					2				4			1	2			1
L2 cache associativity (ways)			16							8					16					8
APU total L3 cache (MiB)			N/A							4		8			N/A					4
APU L3 cache associativity (ways)			N/A							16										16
L3 cache scheme			Victim			N/A				Victim										Victim
Max stock DRAM support			DDR3-1866		DDR3-2133			DDR3-2133, DDR4-2400	DDR4-2400	DDR4-2933		DDR4-3200, LPDDR4-4266			DDR3L-1333	DDR3L-1600	DDR3L-1866		DDR3-1866, DDR4-2400	DDR4-2400
Max DRAM channels per APU			2												1					2
Max stock DRAM bandwidth (GB/s) per APU			29.866		34.132			38.400		46.932		68.256	?		10.666	12.800	14.933		19.200	38.400
GPU microarchitecture			TeraScale 2 (VLIW5)	TeraScale 3 (VLIW4)		GCN 2nd gen		GCN 3rd gen		GCN 5th gen^[4]					TeraScale 2 (VLIW5)	GCN 2nd gen			GCN 3rd gen^[4]	GCN 5th gen
GPU instruction set			TeraScale instruction set			GCN instruction set									TeraScale instruction set	GCN instruction set
Max stock GPU base clock (MHz)			600	800	844	866		1108		1250	1400	2100	2000		538	600	?	847	900	1200
Max stock GPU base GFLOPS^[f]			480	614.4	648.1	886.7		1134.5		1760	1971.2	2150.4	?		86	?	?	?	345.6	460.8
3D engine^[g]			Up to 400:20:8	Up to 384:24:6		Up to 512:32:8				Up to 704:44:16^[5]		Up to 512:32:8			80:8:4	128:8:4			Up to 192:?:?	Up to 192:?:?
3D engine^[g]			IOMMUv1			IOMMUv2								IOMMUv1			?		IOMMUv2
Video decoder			UVD 3.0			UVD 4.2		UVD 6.0		VCN 1.0^[6]		VCN 2.0^[7]		UVD 3.0	UVD 4.0	UVD 4.2	UVD 6.0	UVD 6.3	VCN 1.0
Video encoder			N/A	VCE 1.0		VCE 2.0		VCE 3.1		VCN 1.0^[6]		VCN 2.0^[7]		N/A	VCE 2.0		VCE 3.1		VCN 1.0
AMD Fluid Motion
GPU power saving			PowerPlay	PowerTune										PowerPlay	PowerTune^[8]
TrueAudio			N/A			^[9]								N/A
FreeSync			N/A			1 2								N/A	1 2
HDCP^[h]			?			1.4				1.4 2.2				?	1.4				1.4 2.2
PlayReady^[h]			N/A							3.0 not yet				N/A					3.0 not yet
Supported displays^[i]			2–3	2–4				3		3 (desktop) 4 (mobile, embedded)		4		2				3	4
`/drm/radeon`^[j]^[11]^[12]									N/A									N/A
`/drm/amdgpu`^[j]^[13]			N/A			^[14]								N/A	^[14]

^ For FM2+ Excavator models: A8-7680, A6-7480 & Athlon X4 845.
^ A PC would be one node.
^ An APU combines a CPU and a GPU. Both have cores.
^ Requires firmware support.
^ No SSE4. No SSSE3.
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
^ Unified shaders : texture mapping units : render output units
^ a b To play protected video content, it also requires card, operating system, driver, and application support. A compatible HDCP display is also needed for this. HDCP is mandatory for the output of certain audio formats, placing additional constraints on the multimedia setup.
^ To feed more than two displays, the additional panels must have native DisplayPort support.^[10] Alternatively active DisplayPort-to-DVI/HDMI/VGA adapters can be employed.
^ a b DRM (Direct Rendering Manager) is a component of the Linux kernel. Support in this table refers to the most current version.

Graphics API overview[edit]

The following table shows the graphics and compute APIs support across AMD GPU micro-architectures. Note that this table include micro-architectures not used in the APUs, and a branding series might include older generation chips.

Chip series	Microarchitecture	Fab	Supported APIs					AMD support	Year introduced	Introduced with
			Rendering			Computing
			Vulkan^[15]	OpenGL^[16]	Direct3D	HSA	OpenCL
Wonder	Fixed-pipeline^[a]	1000 nm 800 nm	N/A	N/A	N/A	N/A	N/A	Ended	1986
Mach		800 nm 600 nm			N/A				1991
3D Rage		500 nm			5.0				1996	3D Rage
Rage Pro		350 nm		1.1	6.0				1997	Rage Pro
Rage 128		250 nm		1.2	6.0				1998	Rage 128 GL/VR
R100		180 nm 150 nm		1.3	7.0				2000	Radeon
R200	Programmable pixel & vertex pipelines	150 nm		1.3	8.1				2001	Radeon 8500
R300		150 nm 130 nm 110 nm		2.0^[b]	9.0 11 (FL 9_2)				2002	Radeon 9700
R420		130 nm 110 nm			9.0b 11 (FL 9_2)				2004	Radeon X800
R520		90 nm 80 nm			9.0c 11 (FL 9_3)				2005	Radeon X1800
R600	TeraScale 1	80 nm 65 nm		3.3	10.0 11 (FL 10_0)		ATI Stream		2007	Radeon HD 2900 XT
RV670		55 nm			10.1 11 (FL 10_1)		ATI Stream APP^[17]		2007	Radeon HD 3850/3870
RV770		55 nm 40 nm			10.1 11 (FL 10_1)		1.0		2008	Radeon HD 4850/4870
Evergreen	TeraScale 2	40 nm		4.5 (Linux 4.2) ^[18]^[19]^[20]^[c]	11 (FL 11_0)		1.2		2009	Radeon HD 5850/5870
Northern Islands	TeraScale 2 TeraScale 3	40 nm		4.5 (Linux 4.2) ^[18]^[19]^[20]^[c]	11 (FL 11_0)		1.2		2010	Radeon HD 6850/6870 Radeon HD 6950/6970
Southern Islands	GCN 1st gen	28 nm	1.0	4.6	11 (FL 11_1) 12 (FL11_1)		1.2 2.0 possible	Current	2012	Radeon HD 7950/7970
Sea Islands	GCN 2nd gen		1.2		11 (FL 12_0) 12 (FL 12_0)		2.0 (1.2 in MacOS, Linux) 2.1 Beta in Linux ROCm 2.2 possible		2013	Radeon HD 7790
Volcanic Islands	GCN 3rd gen								2014	Radeon R9 285
Arctic Islands	GCN 4th gen	14 nm							2016	Radeon RX 480
Vega	GCN 5th gen	14 nm 7 nm			11 (FL 12_1) 12 (FL 12_1)				2017	Radeon Vega Frontier Edition
Navi	RDNA 1st gen	7 nm							2019	Radeon RX 5700 (XT)
Navi 2X	RDNA 2nd gen	7 nm							2020	Radeon RX 6800 (XT)

^ Radeon 7000 Series has programmable pixel shaders, but do not fully comply with DirectX 8 or Pixel Shader 1.0. See article on R100's pixel shaders.
^ These series do not fully comply with OpenGL 2+ as the hardware does not support all types of non-power-of-two (NPOT) textures.
^ OpenGL 4+ compliance requires supporting FP64 shaders and these are emulated on some TeraScale chips using 32-bit hardware.

^[21]^[22]^[23]

Desktop APUs[edit]

Lynx: "Llano" (2011)[edit]

Socket FM1
CPU: K10 (or Husky or K10.5) with no L3 cache cores with an upgraded architecture known as Stars
- L1 Cache: 64 KB Data per core and 64 KB Instructions per core
MMX, Enhanced 3DNow!, SSE, SSE2, SSE3, SSE4a, ABM, NX bit, AMD64, Cool'n'Quiet, AMD-V
GPU: TeraScale 2 (Evergreen); all A and E series models feature Redwood-class integrated graphics on die (BeaverCreek for the dual-core variants and WinterPark for the quad-core variants). Sempron and Athlon models exclude integrated graphics.^[24]
List of embedded GPU's
Support for up to four DIMMs of up to DDR3-1866 memory
Fabrication 32 nm on GlobalFoundries SOI process; Die size: 228 mm², with 1.178 billion transistors^[25]^[26]
5 GT/s UMI
Integrated PCIe 2.0 controller
Select models support Turbo Core technology for faster CPU operation when the thermal specification permits
Select models support Hybrid Graphics technology to assist a discrete Radeon HD 6450, 6570, or 6670 discrete graphics card. This is similar to the current Hybrid CrossFireX technology available in the AMD 700 and 800 chipset series

Model	Released	Fab	Stepping	CPU						GPU				DDR3 Memory support	TDP (W)	Box Number	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2	L3	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
Sempron X2 198	2012 Q1	32nm	B0	2 (2)	2.5	N/A	64KB inst. 64KB data per core	2× 512KB	N/A	N/A				1600	65	SD198XOJGXBOX	SD198XOJZ22GX
Athlon II X2 221	2012 Q1			2 (2)	2.8			2× 512KB						1600		AD221XOJGXBOX	AD221XOJZ22GX
Athlon II X4 631 (65 W)	2012			4 (4)	2.6			4× 1MB						1866		AD631XOJGXBOX	AD631XOJZ43GX
Athlon II X4 631 (100 W)	2011 8/15				2.6										100	AD631XOJGXBOX	AD631XWNZ43GX
Athlon II X4 638	2012 2/8				2.7										65	AD638XOJGXBOX	AD638XOJZ43GX
Athlon II X4 641	2012 2/8				2.8										100	AD641XWNGXBOX	AD641XWNZ43GX
Athlon II X4 651	2011 11/14				3.0											AD651XWNGXBOX	AD651XWNZ43GX
Athlon II X4 651K	2012 Q1				3.0											AD651KWNGXBOX	AD651KWNZ43GX
E2-3200	2011 Q4			2 (2)	2.4			2× 512KB		HD 6370D	160:8:4	443	141.7	1600	65	ED3200OJGXBOX	ED3200OJZ22GX ED3200OJZ22HX
A4-3300	2011 9/7				2.5					HD 6410D		443	141.7			AD3300OJGXBOX AD3300OJHXBOX	AD33000OJZ22GX AD33000OJZ22HX
A4-3400	2011 9/7				2.7							600	192			AD3400OJGXBOX AD3400OJHXBOX	AD3400OJZ22GX AD3400OJZ22HX
A4-3420	2011 12/20				2.8							600	192			N/A	AD3420OJZ22HX
A6-3500	2011 8/17			3 (3)	2.1	2.4		3× 1MB		HD 6530D	320:16:8	443	283.5	1866		AD3500OJGXBOX	AD3500OJZ33GX
A6-3600	2011 8/17			4 (4)	2.1	2.4		4× 1MB								AD3600OJGXBOX	AD3600OJZ43GX
A6-3620	2011 12/20				2.2	2.5										AD3620OJGXBOX	AD3620OJZ43GX
A6-3650	2011 6/30				2.6	N/A									100	AD3650WNZ43GX
A6-3670K⁴	2011 12/20				2.7	N/A									100	AD3670WNGXBOX	AD3670WNZ43GX
A8-3800	2011 8/17				2.4	2.7				HD 6550D	400:20:8	600	480		65	AD3800OJZ43GX
A8-3820	2011 12/20				2.5	2.8									65	AD3820OJGXBOX	AD3820OJZ43GX
A8-3850	2011 6/30				2.9	N/A									100	AD3850WNGXBOX	AD3850WNZ43GX
A8-3870K⁴	2011 12/20				3.0	N/A									100	AD3870WNGXBOX	AD3870WNZ43GX

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Virgo: "Trinity" (2012)[edit]

Fabrication 32 nm on GlobalFoundries SOI process
Socket FM2
CPU: Piledriver
- L1 Cache: 16 KB Data per core and 64 KB Instructions per module
GPU TeraScale 3 (VLIW4)
Die Size: 246 mm², 1.303 Billion transistors^[28]
Support for up to four DIMMs of up to DDR3-1866 memory
5 GT/s UMI
GPU (based on VLIW4 architecture) instruction support: DirectX 11, Opengl 4.2, DirectCompute, Pixel Shader 5.0, Blu-ray 3D, OpenCL 1.2, AMD Stream, UVD3
Integrated PCIe 2.0 controller, and Turbo Core technology for faster CPU/GPU operation when the thermal specification permits
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AMD64, AMD-V, AES, CLMUL, AVX, XOP, FMA3, FMA4, F16C,^[29] ABM, BMI1, TBM
Sempron and Athlon models exclude integrated graphics
Select models support Hybrid Graphics technology to assist a Radeon HD 7350, 7450, 7470, 7550, 7570, 7670 discrete graphics card.^[30]^[31] However, it has been found that this does not always improve 3D accelerated graphics performance.^[32]^[33]

Model number	Released	Fab	Step.	CPU					GPU				DDR3 Mem.	TDP (W)	Box Number	Part number^[34]
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	GFLOPS^[b]
Sempron X2 240^[35]		32nm	TN-A1	[1]2	2.9	3.3	64KB inst. per module 16KB data per core	1	N/A				1600	65		SD240XOKA23HJ
Athlon X2 340^[36]	2012/10			[1]2	3.2	3.6		1					1600	65		AD340XOKA23HJ

Athlon X4 730	2012 10/1			[2]4	2.8	3.2		2× 2MB					1866	65		AD730XOKA44HJ
Athlon X4 740	2012/10				3.2	3.7								65	AD740XOKHJBOX	AD740XOKA44HJ
Athlon X4 750K	2012/10				3.4	4.0								100	AD750KWOHJBOX	AD750KWOA44HJ

FirePro A300	2012 8/7				3.4	4.0			FirePro	384:24:8 6 CU	760	583.6		65		AWA300OKA44HJ
FirePro A320	2012 8/7				3.8	4.2			FirePro	384:24:8 6 CU	800	614.4		100		AWA320WOA44HJ

A4-5300	2012 10/1			[1]2	3.4	3.6		1	HD 7480D	128:8:4 2 CU	723	185	1600	65	AD5300OKHJBOX	AD5300OKA23HJ
A4-5300B	2012/10				3.4	3.6			HD 7480D	128:8:4 2 CU	723	185	1600			AD530BOKA23HJ

A6-5400K	2012 10/1				3.6	3.8			HD 7540D	192:12:4 3 CU	760	291.8	1866		AD540KOKHJBOX	AD540KOKA23HJ
A6-5400B	2012/10				3.6	3.8			HD 7540D	192:12:4 3 CU	760	291.8	1866			AD540BOKA23HJ

A8-5500	2012 10/1			[2]4	3.2	3.7		2× 2MB	HD 7560D	256:16:8 4 CU	760	389.1	1866	65	AD5500OKHJBOX	AD5500OKA44HJ
A8-5500B	2012/10				3.2	3.7								65		AD550BOKA44HJ
A8-5600K	2012 10/1				3.6	3.9								100	AD560KWOHJBOX	AD560KWOA44HJ

A10-5700	2012 10/1				3.4	4.0			HD 7660D	384:24:8 6 CU	760	583.6		65	AD5700OKHJBOX	AD5700OKA44HJ
A10-5800K	2012 10/1				3.8	4.2					800	614.4		100	AD580KWOHJBOX	AD580KWOA44HJ
A10-5800B	2012/10				3.8	4.2					800	614.4		100		AD580BWOA44HJ

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Richland" (2013)[edit]

Fabrication 32 nm on GlobalFoundries SOI process
Socket FM2
Two or four CPU cores based on the Piledriver microarchitecture
- Die Size: 246 mm², 1.303 Billion transistors^[37]
- L1 Cache: 16 KB Data per core and 64 KB Instructions per module
- MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AMD64, AMD-V, AES, AVX, AVX1.1, XOP, FMA3, FMA4, F16C, ABM, BMI1, TBM, Turbo Core 3.0, NX bit, PowerNow!
GPU
- TeraScale 3 architecture
- HD Media Accelerator, AMD Hybrid Graphics

Model	Released	Fab	Step.	CPU					GPU				DDR3 Mem.	TDP (W)	Box Number	Part number
				[Modules/FPUs] Cores/threads	Freq. (GHz)		Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
					Base	Turbo	Cache^[a]
					Base	Turbo	L1	L2 (MB)
Sempron X2 250^[35]		32nm	RL-A1	[1]2	3.2	3.6	64KB inst. per module 16KB data per core	1	N/A					65		SD250XOKA23HL

Athlon X2 350^[38]					3.5	3.9							1866	65		AD350XOKA23HL
Athlon X2 370K	Jun 2013				4.0	4.2								65	AD370KOKHLBOX	AD370KOKA23HL

Athlon X4 750	Oct 2013			[2]4	3.4	4.0		2× 2MB						65		AD750XOKA44HL
Athlon X4 760K	Jun 2013				3.8	4.1								100	AD760KWOHLBOX	AD760KWOA44HL

FX-670K^[39]	Mar 2014 (OEM)				3.7	4.3								65		FD670KOKA44HL

A4-4000	May 2013			[1]2	3.0	3.2		1	7480D	128:8:4 2 CU	720	184.3	1333	65	AD4000OKHLBOX	AD4000OKA23HL
A4-4020	Jan 2014				3.2	3.4			7480D		720	184.3	1333		AD4020OKHLBOX	AD4000OKA23HL
A4-6300	July 2013				3.7	3.9			8370D		760	194.5	1600		AD6300OKHLBOX	AD6300OKA23HL
A4-6300B	July 2013				3.7	3.9										AD630BOKA23HL
A4-6320	Dec 2013				3.8	4.0									AD6320OKHLBOX	AD6320OKA23HL
A4-6320B	Mar 2014															AD632BOKA23HL
A4-7300	Aug 2014								8470D	192:12:4 3 CU	800	307.2				AD7300OKA23HL
A4 PRO-7300B	Aug 2014															AD730BOKA23HL

A6-6400B	2013 6/4				3.9	4.1							1866			AD640BOKA23HL
A6-6400K	2013 6/4				3.9	4.1									AD640KOKHLBOX	AD640KOKA23HL
A6-6420B	Jan 2014				4.0	4.2										AD642BOKA23HL
A6-6420K	Jan 2014				4.0	4.2									AD642KOKHLBOX	AD642KOKA23HL

A8-6500T	Sep 18, 2013			[2]4	2.1	3.1		2× 2MB	8550D	256:16:8 4 CU	720	368.6	1866	45	AD650TYHHLBOX	AD650TYHA44HL
A8-6500	2013 6/4				3.5	4.1			8570D		800	409.6		65	AD6500OKHLBOX	AD6500OKA44HL
A8-6500B					3.5	4.1					800	409.6		65		AD650BOKA44HL
A8-6600K					3.9	4.2					844	432.1		100	AD660KWOHLBOX	AD660KWOA44HL

A10-6700T	Sep 18, 2013				2.5	3.5			8650D	384:24:8 4 CU	720	552.9	1866	45	AD670TYHHLBOX	AD670TYHA44HL
A10-6700	2013 6/4				3.7	4.3			8670D		844	648.1		65	AD6700OKHLBOX	AD6700OKA44HL
A10-6790B	Oct 29, 2013				4.0									100	AD679KWOHLBOX	AD679KWOA44HL
A10-6790K	Oct 28, 2013				4.0											AD679BWOA44HL
A10-6800K	2013 6/4				4.1	4.4							2133		AD680KWOHLBOX	AD680KWOA44HL
A10-6800B	2013 6/4				4.1	4.4							2133			AD680BWOA44HL

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Kabini" (2013, SoC)[edit]

Fabrication 28 nm by GlobalFoundries
Socket AM1, aka Socket FS1b (AM1 platform)
2 to 4 CPU Cores (Jaguar (microarchitecture))
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AMD64, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
SoC with integrated memory, PCIe, 2× USB 3.0, 6× USB 2.0, Gigabit Ethernet, and 2× SATA III (6 Gb/s) controllers
GPU based on Graphics Core Next (GCN)

Model	Released	Fab	Step.	CPU					GPU				DDR3 Memory support	TDP (W)	Box Number	Part number
				Cores (threads)	Clock (GHz)	Cache^[a]			Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads)	Clock (GHz)	L1	L2 (MB)	L3	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
Athlon X4 530	?	28nm	A1	4 (4)	2	32 KB inst. 32 KB data per core	2	N/A	N/A				?	25		AD530XJAH44HM
Athlon X4 550	?			4 (4)	2.2		2		N/A				?			AD550XJAH44HM
Sempron 2650	April 9, 2014			2 (2)	1.45		1		R3 (HD 8240)	128:8:4 2 CU	400	102.4	1333 (Single channel only)		SD2650JAHMBOX	SD2650JAH23HM
Sempron 3850				4 (4)	1.30		2		R3 (HD 8280)		450	115.2	1600 (Single channel only)		SD3850JAHMBOX	SD3850JAH44HM
Athlon 5150					1.60				R3 (HD 8400)		600	153.6			AD5150JAHMBOX	AD5150JAH44HM
Athlon 5350					2.05										AD5350JAHMBOX	AD5350JAH44HM
Athlon 5370	Feb, 2016				2.20											AD5370JAH44HM

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Kaveri" (2014) & "Godavari" (2015)[edit]

Fabrication 28 nm by GlobalFoundries.
Socket FM2+,^[40] support for PCIe 3.0.
Two or four CPU cores based on the Steamroller microarchitecture.
- Kaveri refresh models have codename Godavari.^[41]
Die Size: 245 mm², 2.41 Billion transistors.^[42]
L1 Cache: 16 KB Data per core and 96 KB Instructions per module.
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, XOP, FMA3, FMA4, F16C, ABM, BMI1, TBM, Turbo Core
Three to eight Compute Units (CUs) based on GCN 2nd gen microarchitecture;^[43] 1 Compute Unit (CU) consists of 64 Unified Shader Processors : 4 Texture Mapping Units (TMUs) : 1 Render Output Unit (ROPs).
Heterogeneous System Architecture-enabled zero-copy through pointer passing.
SIP blocks: Unified Video Decoder, Video Coding Engine, TrueAudio.^[44]
Dual-channel (2× 64 Bit) DDR3 memory controller.
Integrated custom ARM Cortex-A5 co-processor^[45] with TrustZone Security Extensions^[46] in select APU models, except the Performance APU models.^[47]
Select models support Hybrid Graphics technology by using a Radeon R7 240 or R7 250 discrete graphics card.^[48]
Display controller: AMD Eyefinity 2, 4K Ultra HD support, DisplayPort 1.2 Support.^[49]

Model	Release date	Fab	Step.	CPU					GPU				DDR3 Memory support	TDP (W)	Box Number	Part number
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	GFLOPS^[b]
Athlon X2 450^[38]	Jul 31, 2014	28nm	KV-A1	[1]2	3.5	3.9	96 KB inst. per module 16 KB data per core	1	N/A				1866	65		AD450XYBI23JA

Athlon X4 830	Q1 2018			[2]4	3.0	3.4		2× 2MB					2133			AD830XYBI44JA
Athlon X4 840^[38]	Aug 2014				3.1	3.8									AD840XYBJABOX	AD840XYBI44JA
Athlon X4 850	Q2 2015		GV-A1		3.2											AD835XACI43KA
Athlon X4 860K	Aug 2014		KV-A1		3.7	4.0								95	AD860KXBJABOX AD860KWOHLBOX AD860KXBJASBX	AD860KXBI44JA
Athlon X4 870K	Dec 2015		GV-A1		3.9	4.1									AD870KXBJCSBX	AD870KXBI44JC
Athlon X4 880K	Mar 1, 2016		GV-A1		4.0	4.2									AD880KXBJCSBX

FX-770K^[50]	Dec 2014		KV-A1		3.5	3.9							2133	65		FD770KYBI44JA

A4 PRO-7350B	Jul 31, 2014			[1]2	3.4	3.8		1	R5	192:12:8 3 CU	514	197.3	1866	65		AD735BYBI23JA
PRO A4-8350B	Sep 29, 2015				3.5	3.9				256:16:8 4 CU	757	387.5				AD835BYBI23JC

A6-7400K	Jul 31, 2014				3.5	3.9					756	387			AD740KYBJABOX	AD740KYBI23JA
A6 PRO-7400B	Jul 31, 2014				3.5	3.9					756	387				AD740BYBI23JA
A6-7470K	Feb 2, 2016		GV-A1		3.7	4.0					800	409.6	2133		AD747KYBJCBOX	AD747KYBI23JC
PRO A6-8550B	Sep 29, 2015		GV-A1		3.7	4.0					800	409.6	2133			AD855BYBI23JC

A8-7500^[51]^[52]	2014/?		KV-A1	[2]4	3.0	3.7		2× 2MB	R7	384:24:8 6 CU	720	552.9	2133	65		AD7500YBI44JA
A8-7600	Jul 31, 2014				3.1	3.8									AD7600YBJABOX	AD7600YBI44JA
A8 PRO-7600B	Jul 31, 2014				3.1											AD760BYBI44JA
A8-7650K	Jan 7, 2015				3.3									95	AD765KXBJABOX AD765KXBJASBX	AD765KXBI44JA
A8-7670K	Jul 20, 2015		GV-A1		3.6	3.9					757	581.3		95	AD767KXBJCSBX AD767KXBJCBOX	AD767KXBI44JC
PRO A8-8650B	Sep 29, 2015		GV-A1		3.2	3.9					757	581.3		65		AD865BYBI44JC

A10-7700K	Jan 14, 2014		KV-A1		3.4	3.8					720	552.9		95	AD770KXBJABOX	AD770KXBI44JA
A10-7800	Jul 31, 2014				3.5	3.9				512:32:8 8 CU		737.2		65	AD7800YBJABOX	AD7800YBI44JA
A10 PRO-7800B	Jul 31, 2014				3.5	3.9								65		AD780BYBI44JA
A10-7850K	Jan 14, 2014				3.7	4.0								95	AD785KXBJABOX	AD785KXBI44JA
A10 PRO-7850B	Jul 31, 2014				3.7									95		AD785BXBI44JA
A10-7860K	Feb 2, 2016		GV-A1		3.6						757	775.1		65	AD786KYBJABOX AD786KYBJCSBX	AD786KYBI44JC
A10-7870K	May 28, 2015				3.9	4.1					866	886.7		95	AD787KXDJCBOX AD787KXDJCSBX	AD787KXDI44JC
A10-7890K	Mar 1, 2016				4.1	4.3					866	886.7		95	AD789KXDJCHBX
PRO A10-8750B	Sep 29, 2015				3.6	4.0					757	775.1		65		AD875BYBI44JC
PRO A10-8850B	Sep 29, 2015				3.9	4.1					800	819.2		95		AD885BXBI44JC

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Carrizo" (2016)[edit]

Fabrication 28 nm by GlobalFoundries
Socket FM2+, AM4, support for PCIe 3.0
Two or four CPU cores based on the Excavator microarchitecture
Die Size: 250.04 mm², 3.1 Billion transistors^[53]
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND, Turbo Core
Dual-channel DDR3 or DDR4 memory controller
Third Generation GCN based GPU
Integrated custom ARM Cortex-A5 co-processor^[45] with TrustZone Security Extensions^[46]^[47]

Model	Released	Fab	Step.	Socket	CPU					GPU				DDR3 Memory support	TDP (W)	Box Number^[a]	Part number
					[Modules/FPUs] Cores/threads	Freq.(GHz)		Cache^[b]		Model	Config	Clock (MHz)	GFLOPS^[c]
						Base	Turbo	Cache^[b]
						Base	Turbo	L1	L2 (MB)
Athlon X4 835	?	28nm	CZ-A1	FM2+	[2]4	3.1		96 KB inst. per module 32 KB data per core	2× 1MB	N/A				2133	65		AD835XACI43KA
Athlon X4 845	Feb 2, 2016				[2]4	3.5	3.8		2× 1MB	N/A						AD845XYBJCSBX AD845XACKASBX	AD845XACI43KA
A6-7480^[54]	Oct 2018				[1]2				1	R5	384:24:8 6 CU	900	691.2			AD7480ACABBOX	AD7480ACI23AB
A8-7680^[55]	Oct 2018				[2]4				2× 1MB	R7	384:24:8 6 CU	900	691.2			AD7680ACABBOX	AD7680ACI43AB

PRO A6-8570E	Oct 2016			AM4	[1]2	3.0	3.4		1	R5	256:16:4 4 CU	800	409.6	2400	35		AD857BAHM23AB
PRO A6-8570					[1]2	3.5	3.8		1	R5	384:24:6 6 CU	1029	790.2		65		AD857BAGM23AB
PRO A10-8770E					[2]4	2.8	3.5		2× 1MB	R7		847	650.4		35		AD877BAHM44AB
PRO A10-8770						3.5	3.8					1029	790.2		65		AD877BAGM44AB
PRO A12-8870E						2.9	3.8				512:32:8 8 CU	900	921.6		35		AD887BAHM44AB
PRO A12-8870						3.7	4.2				512:32:8 8 CU	1108	1134.5		65		AD887BAUM44AB

^ With cooler if available.
^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Bristol Ridge" (2016)[edit]

Fabrication 28 nm by GlobalFoundries
Socket AM4, support for PCIe 3.0
Two or four "Excavator+" CPU cores
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND, Turbo Core
Dual-channel DDR4 memory controller
PCI Express 3.0 x8 (No Bifurcation support, requires a PCI-e switch for any configuration other than x8)
PCI Express 3.0 x4 as link to optional external chipset
4x USB 3.1 Gen 1
Storage: 2x SATA and 2x NVMe or 2x PCI Express
Third Generation GCN based GPU^[56] with hybrid VP9 decoding

Model	Released	Fab	Stepping	CPU						GPU				DDR4 Memory support	TDP (W)	Stock Cooler (box)^[a]	Box Number^[b]	Part number
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[c]			Model	Config	Clock (MHz)	Processing power (GFLOPS)^[d]
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2	L3	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[d]
Athlon X4 940	July 27, 2017	28nm	BR-A1	[2]4	3.2	3.6	96 KB inst. per module 32 KB data per core	2× 1 MB	N/A	N/A				2400	65	Near-Silent 65 W	AD940XAGABBOX	AD940XAGM44AB
Athlon X4 950					3.5	3.8											AD950XAGABBOX	AD950XAGM44AB
Athlon X4 970					3.8	4.0											AD970XAUABBOX	AD970XAUM44AB

A6-9500E	September 5, 2016			[1]2	3.0	3.4		1× 1 MB		R5	256:16:4 4 CU	800	409.6		35	Near-Silent 65 W	AD9500AHABBOX	AD9500AHM23AB
PRO A6-9500E	October 3, 2016				3.0	3.4					256:16:4 4 CU	800	409.6		35	?	?	?
A6-9400	March 16, 2019				3.4	3.7					192:12:4 3 CU	720	276.4		65	Z7LH01R201	AD9400AGABBOX
A6-9500	September 5, 2016				3.5	3.8					384:24:6 6 CU	1029	790.2			Near-Silent 65 W	AD9500AGABBOX	AD9500AGM23AB
PRO A6-9500	October 3, 2016				3.5	3.8					384:24:6 6 CU	1029	790.2			?	?	?
A6-9550	July 27, 2017				3.8	4.0					256:16:4 4 CU	800	409.6			Near-Silent 65 W	AD9550AGABBOX	D9550AGM23AB

A8-9600	September 5, 2016			[2]4	3.1	3.4		2× 1 MB		R7	384:24:6 6 CU	900	691.2		65	Near-Silent 65 W	AD9600AGABBOX	AD9600AGM44AB
PRO A8-9600	October 3, 2016				3.1	3.4						900	691.2		65	?	?	?

A10-9700E	September 5, 2016				3.0	3.5						847	650.4		35	Near-Silent 65 W	AD9700AHABBOX	AD9700AHM44AB
PRO A10-9700E	October 3, 2016				3.0	3.5						847	650.4		35	?	?	?
A10-9700	September 5, 2016				3.5	3.8						1029	790.2		65	Near-Silent 65 W	AD9700AGABBOX	D9700AGM44AB
PRO A10-9700	October 3, 2016				3.5	3.8						1029	790.2		65	?	?	?

A12-9800E	September 5, 2016				3.1	3.8					512:32:8^[57] 8 CU	900	921.6		35	Near-Silent 65 W	AD9800AHABBOX	AD9800AUM44AB
PRO A12-9800E	October 3, 2016				3.1	3.8						900	921.6		35	?	?	?
A12-9800	September 5, 2016				3.8	4.2						1108	1134.5		65	Near-Silent 65 W	AD9800AUM44AB
PRO A12-9800	October 3, 2016				3.8	4.2						1108	1134.5		65	?	?	?

^ A box without cooler might also be available (WOF).
^ With cooler if available.
^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Raven Ridge" (2018)[edit]

Fabrication 14 nm by GlobalFoundries
Transistors: 4.94 billion
Die size: 210 mm²
Socket AM4
Zen CPU cores
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Dual-channel DDR4 memory controller
Fifth generation GCN based GPU
Video Core Next (VCN) 1.0

Model	Release Date & Price	Process	CPU						GPU				Socket	PCIe lanes	Memorysupport	TDP	Stock cooler (box)^[a]	Box Number	Part Number
			Cores (threads)	Clock rate (GHz)		Cache^[i]			Model	Config^[ii]	Clock	Processing power (GFLOPS)^[iii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[ii]	Clock	Processing power (GFLOPS)^[iii]
Athlon 200GE^[59]	September 6, 2018 US $55	GloFo 14LP	2 (4)	3.2	N/A	64 KB inst. 32 KB data per core	512 KB per core	4 MB	Vega 3	192:12:4 3 CU	1000 MHz	384	AM4	16 (8+4+4)	DDR4-2666 dual-channel	35 W	Basic Stock Cooler	YD200GC6FBBOX	YD200GC6M2OFB YD20GGC6M2OFB
Athlon Pro 200GE^[60]	September 6, 2018 OEM			3.2													N/A	N/A	YD200BC6M2OFB
Athlon 220GE^[61]	December 21, 2018 US $65			3.4													Basic Stock Cooler	YD220GC6FBBOX	YD220GC6M2OFB
Athlon 240GE^[62]	December 21, 2018 US $75			3.5														YD240GC6FBBOX	YD240GC6M2OFB
Athlon 3000G^[63]	November 19, 2019 US $49			3.5							1100 MHz	424.4						YD3000C6FHBOX	YD3000C6M2OFH
Athlon 300GE^[64]	July 7, 2019 OEM			3.4													N/A	N/A	YD30GEC6M2OFH
Athlon Silver 3050GE^[65]	July 21, 2020 OEM			3.4															YD305GC6M2OFH
Ryzen 3 2200GE^[66]	April 19, 2018 OEM		4 (4)	3.2	3.6				Vega 8	512:32:16 8 CU		1126			DDR4-2933 dual-channel				YD2200C6M4MFB
Ryzen 3 Pro 2200GE^[67]	May 10, 2018 OEM			3.2	3.6														YD220BC6M4MFB
Ryzen 3 2200G	February 12, 2018 US $99			3.5	3.7											45–65 W	Wraith Stealth	YD2200C5FBBOX	YD2200C5M4MFB
Ryzen 3 Pro 2200G^[68]	May 10, 2018 OEM			3.5	3.7											45–65 W	N/A	N/A	YD220BC5M4MFB
Ryzen 5 2400GE^[69]	April 19, 2018 OEM		4 (8)	3.2	3.8				RX Vega 11	704:44:16	1250 MHz	1760				35 W			YD2400C6M4MFB
Ryzen 5 Pro 2400GE^[70]	May 10, 2018 OEM			3.2	3.8											35 W			YD240BC6M4MFB
Ryzen 5 2400G^[71]	February 12, 2018^[72]^[73] US $169			3.6	3.9											45–65 W	Wraith Stealth	YD2400C5FBBOX	YD2400C5M4MFB
Ryzen 5 Pro 2400G^[74]	May 10, 2018 OEM			3.6	3.9											45–65 W	N/A	N/A	YD240BC5M4MFB

^ AMD defines 1 kilobyte (KB) as 1024 bytes, and 1 megabyte (MB) as 1024 kilobytes.^[58]
^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Picasso" (2019)[edit]

Fabrication 12 nm by GlobalFoundries
Transistors: 4.94 billion
Die size: 210 mm²
Socket AM4
Zen+ CPU cores
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Dual-channel DDR4 memory controller
Fifth generation GCN based GPU
Video Core Next (VCN) 1.0

Model	Release Date & Price	Fab	CPU						GPU				Socket	PCIe lanes	Memorysupport	TDP
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
Athlon Pro 300GE^[75]	September 30, 2019 OEM	12 nm	2 (4)	3.4	N/A	64 KB inst. 32 KB data per core	512 KB per core	4 MB	Vega 3	192:12:4 3 CU	1.1	424.4	AM4	16 (8+4+4)	DDR4-2667 dual-channel	35 W
Athlon Silver Pro 3125GE^[76]	July 21, 2020 OEM		2 (4)	3.4	N/A										DDR4-2667 dual-channel
Athlon Gold 3150GE^[77]			4 (4)	3.3	3.8										DDR4-2933 dual-channel
Athlon Gold Pro 3150GE^[78]				3.3	3.8
Athlon Gold 3150G^[79]				3.5	3.9											45-65 W
Athlon Gold Pro 3150G^[80]				3.5	3.9											45-65 W
Ryzen 3 3200GE^[81]	July 7, 2019 OEM			3.3	3.8				Vega 8	512:32:16 8 CU	1.2	1228.8				35 W
Ryzen 3 Pro 3200GE^[82]	September 30, 2019 OEM			3.3	3.8						1.2	1228.8				35 W
Ryzen 3 3200G^[83]	July 7, 2019 US $99			3.6	4.0						1.25	1280				45-65 W
Ryzen 3 Pro 3200G^[84]	September 30, 2019 OEM			3.6	4.0						1.25	1280				45-65 W
Ryzen 5 3350GE	July 21, 2020 OEM		4 (8)	3.3	3.9				RX Vega 11	704:44:16 11 CU	1.2					35 W
Ryzen 5 Pro 3350GE^[85]				3.3	3.9						1.2					35 W
Ryzen 5 3350G				3.6	4.0						1.3	1830.4				45-65 W
Ryzen 5 Pro 3350G^[86]				3.6	4.0											45-65 W
Ryzen 5 3400GE^[87]	July 7, 2019 OEM			3.3	4.0											35 W
Ryzen 5 Pro 3400GE^[88]	September 30, 2019 OEM			3.3	4.0											35 W
Ryzen 5 3400G^[89]	July 7, 2019 US $149			3.7	4.2						1.4	1971.2				45-65 W
Ryzen 5 Pro 3400G^[90]	September 30, 2019 OEM			3.7	4.2						1.4	1971.2				45-65 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Renoir" (2020)[edit]

Fabrication 7 nm by TSMC
Socket AM4
Up to eight Zen 2 CPU cores
Dual-channel DDR4 memory controller

Model	Release date and price	Fab	CPU							GPU				Socket	PCIe lanes	Memorysupport	TDP
			Cores (threads)	Core Config^[i]	Clock rate (GHz)		Cache			Model	Config^[ii]	Clock (GHz)	Processing power (GFLOPS)^[iii]
			Cores (threads)	Core Config^[i]	Base	Boost	L1	L2	L3	Model	Config^[ii]	Clock (GHz)	Processing power (GFLOPS)^[iii]
Ryzen 3 4300GE ^[91]	July 21, 2020	TSMC 7FF	4 (8)	1 × 4	3.5	4.0	32 KB inst. 32 KB data per core	512 KB per core	4 MB	Vega 6	384:24:8 6 CU	1.7	1305.6	AM4	24 (16+4+4)	DDR4-3200 dual-channel	35 W
Ryzen 3 Pro 4350GE^[91]					3.5	4.0											35 W
Ryzen 3 4300G^[91]					3.8	4.0											65 W
Ryzen 3 Pro 4350G^[91]					3.8	4.0											65 W
Ryzen 5 4600GE^[91]			6 (12)	2 × 3	3.3	4.2			8 MB 4 MB per CCX	Vega 7	448:28:8 7 CU	1.9	1702.4				35 W
Ryzen 5 Pro 4650GE^[91]					3.3	4.2											35 W
Ryzen 5 4600G^[91]					3.7	4.2											65 W
Ryzen 5 Pro 4650G^[91]					3.7	4.2											65 W
Ryzen 7 4700GE^[91]			8 (16)	2 × 4	3.1	4.3				Vega 8	512:32:8 8 CU	2.0	2048				35 W
Ryzen 7 Pro 4750GE^[91]					3.1	4.3						2.0	2048				35 W
Ryzen 7 4700G^[91]					3.6	4.4						2.1	2150.4				65 W
Ryzen 7 Pro 4750G^[91]					3.6	4.4						2.1	2150.4				65 W

^ Core complexes (CCXs) × cores per CCX
^ Unified shaders : texture mapping units : render output units and compute units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Cezanne" (2021)[edit]

Fabrication 7 nm by TSMC
Socket AM4
Up to eight Zen 3 CPU cores
Dual-channel DDR4 memory controller

Model	Release date and price	Fab	CPU							GPU				Socket	PCIe lanes	Memory support	TDP
			Cores (threads)	Core config^[i]	Clock rate (GHz)		Cache			Model	Config^[ii]	Clock	Processing power (GFLOPS)^[iii]
			Cores (threads)	Core config^[i]	Base	Boost	L1	L2	L3	Model	Config^[ii]	Clock	Processing power (GFLOPS)^[iii]
Ryzen 3 5300GE^[92]	13 April 2021 OEM^[iv]	TSMC 7FF	4 (8)	1 × 4^[93]	3.6	4.2	32 KB inst. 32 KB data per core	512 KB per core	8 MB	AMD Radeon Graphics	384:24:8 6 CU	1700 MHz	1305.6	AM4	24 (16+4+4)^[93]	DDR4-3200 dual-channel	35 W
Ryzen 3 5300G^[94]			4 (8)	1 × 4^[93]	4.0	4.2			8 MB		384:24:8 6 CU	1700 MHz	1305.6				65 W
Ryzen 5 5600GE^[95]			6 (12)	1 x 6	3.4	4.4			16 MB		448:28:8 7 CU	1900 MHz	1702.4				35 W
Ryzen 5 5600G^[96]			6 (12)	1 x 6	3.9	4.4					448:28:8 7 CU	1900 MHz	1702.4				65 W
Ryzen 7 5700GE^[97]			8 (16)	1 x 8	3.2	4.6					512:32:8 8 CU	2000 MHz	2048				35 W
Ryzen 7 5700G^[98]			8 (16)	1 x 8	3.8	4.6					512:32:8 8 CU	2000 MHz	2048				65 W

^ Core Complexes (CCX) × cores per CCX
^ Unified shaders : texture mapping units : render output units and compute units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
^ As of April 2021, retail availability expected later in 2021^[93].

Server APUs[edit]

Opteron X2100-series "Kyoto" (2013)[edit]

Fabrication 28 nm
Socket FT3 (BGA)
4 CPU Cores (Jaguar (microarchitecture))
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
Turbo Dock Technology, C6 and CC6 low power states
GPU based on GCN (Graphics Core Next) architecture

				Cores (threads)	Clock (GHz)	Cache^[b]		Model	Config	Clock (MHz)	Processing power (GFLOPS)^[c]
Model	Released	Fab	Stepp.	CPU				GPU				Memory support	TDP (W)	Part number	Release price (USD)
				Cores (threads)	Clock (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[c]
X1150	May 2013	28nm	KB-A1	4 (4)	2.0	32 KB inst. 32 KB data per core	2	N/A				DDR3	9		$64
X2150	May 2013				1.9			HD 8400	128:8:4	266	28.9	DDR3	22	OX2150IAJ44HM	$99
X2170	September 2016				2.4			HD 8400	128:8:4	600	153.6	DDR3	25	OX2170IXJ44JB

^ A box without cooler might also be available (WOF).
^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Opteron X3000-series "Toronto" (2017)[edit]

Fabrication 28 nm
Socket FP4
Two or Four CPU cores based on the Excavator microarchitecture^[99]^[100]
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND
DDR4 SDRAM
GPU based on 3rd Generation GCN (Graphics Core Next) architecture

				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (GHz)	Processing power (GFLOPS)^[b]
Model	Released	Fab	Step.	CPU					GPU				DDR4 Memory support	TDP (W)	Part number	Release price (USD)
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (GHz)	Processing power (GFLOPS)^[b]
X3216	Q2 2017	28nm	01h	2 (2)	1.6	3.0	96 KB inst. per module 32 KB data per core	1	R5	256:16:4 4 CU	0.8	409.6	1600	12- 15	Unknown	OEM for HP
X3418				4 (4)	1.8	3.2		2	R7	384:24:6 6 CU		614.4	2400	12- 35
X3421				4 (4)	2.1	3.4		2	R7	512:32:8 8 CU		819.2	2400	12- 35

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Mobile APUs[edit]

Sabine: "Llano" (2011)[edit]

Fabrication 32 nm on GlobalFoundries SOI process
Socket FS1
Upgraded Stars (AMD 10h architecture) codenamed Husky CPU cores (K10.5) with no L3 cache, and with Redwood-class integrated graphics on die
L1 Cache: 64 KB Data per core and 64 KB Instructions per core(BeaverCreek for the dual-core variants and WinterPark for the quad-core variants)
Integrated PCIe 2.0 controller
GPU: TeraScale 2
Select models support Turbo Core technology for faster CPU operation when the thermal specification permits
Support for 1.35 V DDR3L-1333 memory, in addition to regular 1.5 V DDR3 memory specified
2.5 GT/s UMI
MMX, Enhanced 3DNow!, SSE, SSE2, SSE3, SSE4a, ABM, NX bit, AMD64, AMD-V
PowerNow!

Model	Released	Fab	Step.	CPU						GPU				DDR3 Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config	Clock (MHz)	GFLOPS^[b]
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2	L3	Model	Config	Clock (MHz)	GFLOPS^[b]
E2-3000M	2011 6/14	32nm	B0	2 (2)	1.8	2.4	64 KB inst. 64 KB data per core	2× 512KB	N/A	HD 6380G	160:8:4	400	128	1333	35	EM3000DDX22GX

A4-3300M	2011 6/14				1.9	2.5		2× 1MB		HD 6480G	240:12:4	444	213.1		35	AM3300DDX23GX
A4-3305M	Dec 7, 2011				1.9			2× 512KB			160:8:4	593	189.7		35	AM3305DDX22GX
A4-3310MX	2011 6/14				2.1			2× 1MB			240:12:4	444	213.1		45	AM3310HLX23GX
A4-3320M	Dec 7, 2011				2.0	2.6									35	AM3320DDX23GX
A4-3330MX	Dec 7, 2011				2.2										45	AM3330HLX23GX
A4-3330MX					2.3			2× 512KB			160:8:4	593	189.7		45	AM3330HLX23HX

A6-3400M	2011 6/14			4 (4)	1.4	2.3		4× 1MB		HD 6520G	320:16:8	400	256		35	AM3400DDX43GX
A6-3410MX	2011 6/14				1.6	2.3								1600	45	AM3410HLX43GX
A6-3420M	Dec 7, 2011				1.5	2.4								1333	35	AM3420DDX43GX
A6-3430MX	Dec 7, 2011				1.7	2.4								1600	45	AM3430HLX43GX

A8-3500M	2011 6/14				1.5	2.4				HD 6620G	400:20:8	444	355.2	1333	35	AM3500DDX43GX
A8-3510MX	2011 6/14				1.8	2.5								1600	45	AM3510HLX43GX
A8-3520M	Dec 7, 2011				1.6	2.5								1333	35	AM3520DDX43GX
A8-3530MX	2011 6/14				1.9	2.6								1600	45	AM3530HLX43GX
A8-3550MX	Dec 7, 2011				2.0	2.7								1600	45	AM3550HLX43GX

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Comal: "Trinity" (2012)[edit]

An AMD A10-4600M APU

Fabrication 32 nm on GlobalFoundries SOI process
Socket FS1r2, FP2
Based on the Piledriver architecture
L1 Cache: 16 KB Data per core and 64 KB Instructions per module
GPU: TeraScale 3 (VLIW4)
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, XOP, FMA3, FMA4, F16C, ABM, BMI1, TBM, Turbo Core
Memory support: 1.35 V DDR3L-1600 memory, in addition to regular 1.5 V DDR3 memory specified (Dual-channel)
2.5 GT/s UMI
Transistors: 1.303 billion
Die size: 246 mm²

Model number	Released	Fab	Step.	Socket	CPU					GPU					DDR3 Memory support	TDP (W)	Part number
					[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
					[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
A4-4355M	Sep 27, 2012	32nm	TN-A1	FP2	[1]2	1.9	2.4	64 KB inst. per module 16 KB data per core	1	HD 7400G	192:12:4 3 CU	327	424	125.5	1333	17	AM4355SHE23HJ
A6-4455M	May 15, 2012				[1]2	2.1	2.8		2	HD 7500G	256:16:8 4 CU	327		167.4		17	AM4455SHE24HJ
A8-4555M	Sep 27, 2012				[2]4	1.6	2.4		2× 2MB	HD 7600G	384:24:8 6 CU	320		245.7		19	AM4555SHE44HJ
A8-4557M^[101]	Mar 2013					1.9	2.8			HD 7000	256:16:8 4 CU	497	655	254.4	(L)1600	35	AM4557DFE44HJ
A10-4655M	May 15, 2012					2.0	2.8			HD 7620G	384:24:8 6 CU	360	496	276.4	1333	25	AM4655SIE44HJ
A10-4657M^[101]	Mar 2013					2.3	3.2			HD 7000	384:24:8 6 CU	497	686	381.6	(L)1600	35	AM4657DFE44HJ

A4-4300M	May 15, 2012			FS1	[1]2	2.5	3.0		1	HD 7420G	128:8:4 2 CU	480	655	122.8	1600		AM4300DEC23HJ
A6-4400M					[1]2	2.7	3.2		1	HD 7520G	192:12:4 3 CU	496	685	190.4			AM4400DEC23HJ
A8-4500M					[2]4	1.9	2.8		2× 2MB	HD 7640G	256:16:8 4 CU			253.9			AM4500DEC44HJ
A10-4600M					[2]4	2.3	3.2		2× 2MB	HD 7660G	384:24:8 6 CU			380.9			AM4600DEC44HJ

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Richland" (2013)[edit]

Fabrication 32 nm on GlobalFoundries SOI process
Socket FS1r2, FP2
Elite Performance APU.^[102]^[103]
CPU: Piledriver architecture
- L1 Cache: 16 KB Data per core and 64 KB Instructions per module
GPU: TeraScale 3 (VLIW4)
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, XOP, FMA3, FMA4, F16C, ABM, BMI1, TBM, Turbo Core

Model number	Released	Fab	Step.	Socket	CPU					GPU					DDR3 Memory support	TDP (W)	Part number
					[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
					[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
A4-5145M	2013/5	32nm	RL-A1	FP2	[1]2	2.0	2.6	64 KB inst. per module 16 KB data per core	1	HD 8310G	128:8:4 2 CU	424	554	108.5	(L)1333	17	AM5145SIE44HL
A6-5345M					[1]2	2.2	2.8		1	HD 8410G	192:12:4 3 CU	450	600	172.8		17	AM5345SIE44HL
A8-5545M					[2]4	1.7	2.7		4	HD 8510G	384:28:8 6 CU	450	554	345.6		19	AM5545SIE44HL
A10-5745M					[2]4	2.1	2.9		4	HD 8610G	384:28:8 6 CU	533	626	409.3		25	AM5745SIE44HL

A4-5150M	2013 Q1			FS1	[1]2	2.7	3.3		1	HD 8350G	128:8:4 2 CU	533	720	136.4	1600	35	AM5150DEC23HL
A6-5350M	2013 Q1					2.9	3.5			HD 8450G	192:12:4 3 CU			204.6	1600		AM5350DEC23HL
A6-5357M	2013/5					2.9	3.5			HD 8450G	192:12:4 3 CU			204.6	(L)1600		AM5357DFE23HL
A8-5550M	2013 Q1				[2]4	2.1	3.1		4	HD 8550G	256:16:8 4 CU	515		263.6	1600		AM5550DEC44HL
A8-5557M	2013/5					2.1	3.1			HD 8550G	256:16:8 4 CU	554		283.6	(L)1600		AM5557DFE44HL
A10-5750M	2013 Q1					2.5	3.5			HD 8650G	384:24:8 6 CU	533		409.3	1866		AM5750DEC44HL
A10-5757M	2013/5					2.5	3.5			HD 8650G	384:24:8 6 CU	600		460.8	(L)1600		AM5757DFE44HL

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Kaveri" (2014)[edit]

Fabrication 28 nm
Socket FP3
Up to 4 Steamroller x86 CPU cores with 4 MB of L2 cache.^[104]
L1 Cache: 16 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, XOP, FMA3, FMA4, F16C, ABM, BMI1, TBM, Turbo Core
Three to eight Compute Units (CUs) based on Graphics Core Next (GCN)^[43] microarchitecture; 1 Compute Unit (CU) consists of 64 Unified Shader Processors : 4 Texture Mapping Units (TMUs) : 1 Render Output Unit (ROPs)
AMD Heterogeneous System Architecture (HSA) 2.0
SIP blocks: Unified Video Decoder, Video Coding Engine, TrueAudio^[44]
Dual-channel (2x64-bit) DDR3 memory controller
Integrated custom ARM Cortex-A5 co-processor^[45] with TrustZone Security Extensions^[46]

Model number	Released	Fab	CPU					GPU					DDR3 Memory support	TDP (W)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
A6-7000	June 2014	28nm	[1]2	2.2	3.0	96 KB inst. per module 16 KB data per core	1	R4	192:12:3 3 CU	494	533	189.6	1333	17	AM7000ECH23JA
A6 Pro-7050B			[1]2	2.2	3.0		1	R4	192:12:3 3 CU	533	N/A	204.6	1600	17	AM705BECH23JA

A8-7100			[2]4	1.8	3.0		2× 2 MB	R5	256:16:4 4 CU	450	514	230.4	1600	20	AM7100ECH44JA
A8 Pro-7150B				1.9	3.2			R5	256:16:4 4 CU	553	N/A	283.1			AM715BECH44JA
A10-7300				1.9	3.2			R6	384:24:8 6 CU	464	533	356.3			AM7300ECH44JA
A10 Pro-7350B				2.1	3.3			R6		533	N/A	424.7			AM735BECH44JA
FX-7500				2.1	3.3			R7		498	553	382.4			FM7500ECH44JA

A8-7200P				2.4	3.3			R5	256:16:4 4 CU	553	626	283.1	1866	35	AM740PDGH44JA
A10-7400P				2.5	3.4			R6	384:24:8 6 CU	576	654	442.3	1866		AM740PDGH44JA
FX-7600P				2.7	3.6			R7	512:32:8 8 CU	600	686	614.4	2133		FM760PDGH44JA

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Carrizo" (2015)[edit]

Fabrication 28 nm
Socket FP4
Up to 4 Excavator x86 CPU cores
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND, Turbo Core
GPU based on Graphics Core Next 1.2

Model number	Released	Fab	CPU					GPU				DDR Memory support	TDP (W)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	GFLOPS^[b]
A6-8500P	June 2015	28nm	[1]2	1.6	3.0	96 KB inst. per module 32 KB data per core	1	R5	256:16:4 4 CU	800	409.6	3)1600	12- 35	AM850PAAY23KA
Pro A6-8500B	June 2015			1.6	3.0							3)1600		AM850BAAY23KA
Pro A6-8530B	Q3 2016			2.3	3.2							4)1866		AM853BADY23AB

A8-8600P	June 2015		[2]4	1.6	3.0		2× 1MB	R6	384:24:8 6 CU	720	552.9	3)2133		AM860PAAY43KA
Pro A8-8600B				1.6	3.0					720	552.9			AM860BAAY43KA
A10-8700P				1.8	3.2					800	614.4			AM870PAAY43KA
Pro A10-8700B				1.8	3.2					800	614.4			AM870BAAY43KA
Pro A10-8730B	Q3 2016			2.4	3.3			R5		720	552.9	4)1866		AM873BADY44AB
A10-8780P	December 2015			2.0	3.3			R8	512:32:8 8 CU			3)?		AM878PAIY43KA
FX-8800P	June 2015			2.1	3.4			R7		800	819.2	4)2133		FM880PAAY43KA
Pro A12-8800B	June 2015			2.1	3.4					800	819.2	4)2133		FM880BAAY43KA
Pro A12-8830B	Q3 2016			2.5	3.4				384:24:8 6 CU	758	582.1	4)1866		AM883BADY44AB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Bristol Ridge" (2016)[edit]

Fabrication 28 nm
Socket FP4^[105]
Two or four "Excavator+" x86 CPU cores
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND, Turbo Core
GPU based on Graphics Core Next 1.2 with VP9 decoding

Model number	Released	Fab	CPU					GPU				DDR4 Memory support	TDP (W)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	GFLOPS^[b]
Pro A6-9500B	October 24, 2016	28nm	[1]2	2.3	3.2	96 KB inst. per module 32 KB data per core	1	R5	256:16:4 4 CU	800	409.6	1866	12- 15

Pro A8-9600B	October 24, 2016		[2]4	2.4	3.3		2× 1 MB	R5	384:24:6 6 CU	720	552.9	1866	12– 15
A10-9600P	June 2016			2.4	3.3					720	552.9			AM960PADY44AB
A10-9620P^[106]	2017 (OEM)			2.5	3.4					758	582.1
Pro A10-9700B	October 24, 2016							R7
A12-9700P	June 2016							R7						AM970PADY44AB
Pro A8-9630B	October 24, 2016			2.6	3.3			R5		800	614.4	2400	25– 45
A10-9630P	June 2016			2.6	3.3			R5		800	614.4			AM963PAEY44AB
Pro A10-9730B	October 24, 2016			2.8	3.5			R7		900	691.2
A12-9730P	June 2016			2.8	3.5			R7		900	691.2			AM973PAEY44AB

Pro A12-9800B	October 24, 2016			2.7	3.6			R7	512:32:8 8 CU	758	776.1	1866	12– 15
FX-9800P A12-9720P^[107]^[108]	June 2016 2017 (OEM)			2.7	3.6					758	776.1	1866	12– 15	FM980PADY44AB ?
Pro A12-9830B	October 24, 2016			3.0	3.7					900	921.6	2400	25– 45
FX-9830P	June 2016			3.0	3.7					900	921.6	2400	25– 45	FM983PAEY44AB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Raven Ridge" (2017)[edit]

Fabrication 14 nm by GlobalFoundries
Transistors: 4.94 billion
Socket FP5
Die size: 210 mm²
Zen CPU cores
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Fifth generation GCN based GPU

Model	Release date	Fab	CPU						GPU				Socket	PCIe lanes	Memory support	TDP
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[i]	Clock	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[i]	Clock	Processing power (GFLOPS)^[ii]
Athlon Pro 200U ^[109]	2019	GloFo 14LP	2 (4)	2.3	3.2	64 KB inst. 32 KB data per core	512 KB per core	4 MB	Vega 3	192:12:4 3 CU	1000 MHz	384	FP5	12 (8+4)	DDR4-2400 dual-channel	12–25 W
Athlon 300U ^[110]	January 6, 2019			2.4	3.3						1000 MHz	384
Ryzen 3 2200U ^[111]	January 8, 2018			2.5	3.4						1100 MHz	422.4
Ryzen 3 3200U ^[112]	January 6, 2019			2.6	3.5						1200 MHz	460.8
Ryzen 3 2300U ^[113]	January 8, 2018		4 (4)	2.0	3.4				Vega 6	384:24:8 6 CU	1100 MHz	844.8
Ryzen 3 Pro 2300U ^[114]	May 15, 2018		4 (4)		3.4				Vega 6	384:24:8 6 CU		844.8
Ryzen 5 2500U ^[115]	October 26, 2017		4 (8)		3.6				Vega 8	512:32:16 8 CU		1126.4
Ryzen 5 Pro 2500U ^[116]	May 15, 2018
Ryzen 5 2600H ^[117]	September 10, 2018			3.2											DDR4-3200 dual-channel	35–54 W
Ryzen 7 2700U ^[118]	October 26, 2017			2.2	3.8				Vega 10	640:40:16 10 CU	1300 MHz	1664			DDR4-2400 dual-channel	12–25 W
Ryzen 7 Pro 2700U ^[119]	May 15, 2018			2.2					Vega 10	640:40:16 10 CU		1664			DDR4-2400 dual-channel	12–25 W
Ryzen 7 2800H ^[120]	September 10, 2018			3.3					Vega 11	704:44:16 11 CU		1830.4			DDR4-3200 dual-channel	35–54 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Picasso" (2019)[edit]

Fabrication 12 nm by GlobalFoundries
Socket FP5
Die size: 210 mm²
Up to four Zen+ CPU cores
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Dual-channel DDR4 memory controller
Fifth generation GCN based GPU

Model	Release date	Fab	CPU						GPU				Memory support	TDP
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[i]	Clock	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[i]	Clock	Processing power (GFLOPS)^[ii]
Ryzen 3 3300U^[121]	January 6, 2019	GloFo 12LP (14LP+)	4 (4)	2.1	3.5	64 KB inst. 32 KB data per core	512 KB per core	4 MB	Vega 6	384:24:8 6 CU^[122]	1200 MHz	921.6	DDR4-2400 dual-channel	15 W
Ryzen 3 PRO 3300U^[123]	April 8, 2019
Ryzen 3 3350U^[124]	January 6, 2019
Ryzen 5 3450U^[125]	June 2020		4 (8)						Vega 8	512:32:16 8 CU^[126]		1228.8
Ryzen 5 3500U^[127]	January 6, 2019				3.7
Ryzen 5 PRO 3500U^[128]	April 8, 2019
Ryzen 5 3500C^[129]	September 22, 2020
Ryzen 5 3550H^[130]	January 6, 2019													35 W
Ryzen 5 3580U^[131]	October 2019								Vega 9	576:36:16 9 CU	1300 MHz	1497.6		15 W
Ryzen 7 3700U^[132]	January 6, 2019			2.3	4.0				Vega 10	640:40:16 10 CU^[133]	1400 MHz	1792.0
Ryzen 7 PRO 3700U^[134]	April 8, 2019
Ryzen 7 3700C^[135]	September 22, 2020
Ryzen 7 3750H^[136]	January 6, 2019													35 W
Ryzen 7 3780U^[137]	October 2019								Vega 11	704:44:16 11 CU		1971.2		15 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Renoir" (2020)[edit]

Fabrication 7 nm by TSMC^[138]^[139]^[140]
Socket FP6
Die size: 156 mm²
9.8 billion transistors on one single 7nm monolithic die^[141]
Up to eight Zen 2 CPU cores
Fifth generation GCN based GPU (7nm Vega)

Model	Release date	Fab	CPU							GPU			Socket	PCIe lanes	Memory support	TDP
			Cores (threads)	Core config^[i]	Clock rate (GHz)		Cache			Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
			Cores (threads)	Core config^[i]	Base	Boost	L1	L2	L3	Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
Ryzen 3 4300U^[142]	March 16, 2020	TSMC 7FF	4 (4)	1 × 4	2.7	3.7	32 KB inst. 32 KB data per core	512 KB per core	4 MB	AMD Radeon Graphics 320:20:8 5 CU	1400 MHz	896	FP6	16 (8+4+4)	DDR4-3200 LPDDR4-4266 dual-channel	10–25 W
Ryzen 3 PRO 4450U^[143]	May 7, 2020		4 (8)	1 × 4	2.5	3.7			4 MB	AMD Radeon Graphics 320:20:8 5 CU	1400 MHz	896
Ryzen 5 4500U^[144]	March 16, 2020		6 (6)	2 × 3	2.3	4.0			8 MB 4 MB per CCX	AMD Radeon Graphics 384:24:8 6 CU	1500 MHz	1152
Ryzen 5 4600U^[145]	March 16, 2020		6 (12)		2.1
Ryzen 5 PRO 4650U^[146]	May 7, 2020
Ryzen 5 4680U^[147]	April 13, 2021									AMD Radeon Graphics 448:28:8 7 CU		1344
Ryzen 5 4600HS^[148]	March 16, 2020				3.0					AMD Radeon Graphics 384:24:8 6 CU		1152				35 W
Ryzen 5 4600H^[149]					3.0					AMD Radeon Graphics 384:24:8 6 CU		1152				35–54 W
Ryzen 7 4700U^[150]			8 (8)	2 × 4	2.0	4.1				AMD Radeon Graphics 448:28:8 7 CU	1600 MHz	1433.6				10–25 W
Ryzen 7 PRO 4750U^[151]	May 7, 2020		8 (16)		1.7	4.1				AMD Radeon Graphics 448:28:8 7 CU	1600 MHz	1433.6
Ryzen 7 4800U^[152]	March 16, 2020				1.8	4.2				AMD Radeon Graphics 512:32:8 8 CU	1750 MHz	1792
Ryzen 7 4980U^[153]	April 13, 2021				2.0	4.4				AMD Radeon Graphics 512:32:8 8 CU	1950 MHz	1996.8
Ryzen 7 4800HS^[154]	March 16, 2020				2.9	4.2				AMD Radeon Graphics 448:28:8 7 CU	1600 MHz	1433.6				35 W
Ryzen 7 4800H^[155]					2.9	4.2				AMD Radeon Graphics 448:28:8 7 CU	1600 MHz	1433.6				35–54 W
Ryzen 9 4900HS^[156]					3	4.3				AMD Radeon Graphics 512:32:8 8 CU	1750 MHz	1792				35 W
Ryzen 9 4900H^[157]					3.3	4.4				AMD Radeon Graphics 512:32:8 8 CU	1750 MHz	1792				35–54 W

^ Core Complexes (CCX) × cores per CCX
^ Unified shaders : texture mapping units : render output units and compute units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Lucienne" (2021)[edit]

Fabrication 7 nm by TSMC
Socket FP6
Die size: 156 mm²
9.8 billion transistors on one single 7nm monolithic die^[141]
Up to eight Zen 2 CPU cores
Fifth generation GCN based GPU (7nm Vega)

Model	Release date	Fab	CPU							GPU			Socket	PCIe lanes	Memory support	TDP
			Cores (threads)	Core config^[i]	Clock rate (GHz)		Cache			Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
			Cores (threads)	Core config^[i]	Base	Boost	L1	L2	L3	Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
Ryzen 3 5300U^[158]	January 12, 2021	TSMC 7FF	4 (8)	1 × 4	2.6	3.8	32 KB inst. 32 KB data per core	512 KB per core	4 MB	AMD Radeon Graphics 320:20:8 6 CU	1500 MHz	1152	FP6	16 (8+4+4)	DDR4-3200 LPDDR4-4266 dual-channel	10–25 W
Ryzen 5 5500U^[159]			6 (12)	2 × 3	2.1	4.0			8 MB 4 MB per CCX	AMD Radeon Graphics 384:24:8 7 CU	1800 MHz	1612.8
Ryzen 7 5700U^[160]			8 (16)	2 × 4	1.8	4.3			8 MB 4 MB per CCX	AMD Radeon Graphics 8 CU	1900 MHz	1945.6

^ Core Complexes (CCX) × cores per CCX
^ Unified shaders : texture mapping units : render output units and compute units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Cezanne" (2021)[edit]

Fabrication 7 nm by TSMC
Socket FP6
Die size: 180 mm²
Up to eight Zen 3 CPU cores
Fifth generation GCN based GPU (7nm Vega)

Model	Release date	Fab	CPU							GPU			Socket	PCIe lanes	Memory support	TDP
			Cores (threads)	Core config^[i]	Clock rate (GHz)		Cache			Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
			Cores (threads)	Core config^[i]	Base	Boost	L1	L2	L3	Model, config^[ii]	Clock	Processing power (GFLOPS)^[iii]
Ryzen 3 5400U^[161]	January 12, 2021	TSMC 7FF	4 (8)	1 × 4	2.6	4.0	32 KB inst. 32 KB data per core	512 KB per core	8 MB	AMD Radeon Graphics 6 CU	1600 MHz	1228.8	FP6	16 (8+4+4)	DDR4-3200 LPDDR4-4266 dual-channel	10–25 W
Ryzen 3 PRO 5450U^[162]	March 16, 2021		4 (8)	1 × 4	2.6	4.0			8 MB	AMD Radeon Graphics 6 CU	1600 MHz	1228.8
Ryzen 5 5600U^[163]	January 12, 2021		6 (12)	1 × 6	2.3	4.2			16 MB	AMD Radeon Graphics 7 CU	1800 MHz	1612.8
Ryzen 5 PRO 5650U^[164]	March 16, 2021				2.3
Ryzen 5 5600H^[165]	January 12, 2021				3.3											35–54 W
Ryzen 5 5600HS^[166]					3.0											35–54 W
Ryzen 7 5800U^[167]			8 (16)	1 × 8	1.9	4.4				AMD Radeon Graphics 8 CU	2000 MHz	2048				10–25 W
Ryzen 7 PRO 5850U^[168]	March 16, 2021				1.9											10–25 W
Ryzen 7 5800H^[169]	January 12, 2021				3.2											35–54 W
Ryzen 7 5800HS^[170]					2.8
Ryzen 9 5900HS^[171]					3.0	4.6					2100 MHz	2150.4
Ryzen 9 5900HX^[172]					3.3	4.6
Ryzen 9 5980HS^[173]					3.0	4.8
Ryzen 9 5980HX^[174]					3.3	4.8

^ Active Core Complexes (CCX) × active cores per CCX.
^ Unified shaders : texture mapping units : render output units and compute units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Ultra-mobile APUs[edit]

Brazos: "Desna", "Ontario", "Zacate" (2011)[edit]

Fabrication 40 nm by TSMC
Socket FT1 (BGA-413)
Based on the Bobcat microarchitecture^[175]
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, ABM, NX bit, AMD64, AMD-V
PowerNow!
DirectX 11 integrated graphics with UVD 3.0
Z-series denote Desna; C-series denote Ontario; and the E-series denotes Zacate
2.50 GT/s UMI (PCIe 1.0 ×4)

Model	Released	Fab	Step.	CPU					GPU					DDR3 Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2	Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
Z-01	Jun 1, 2011	40nm	B0	2 (2)	1.0	N/A	32KB inst. 32KB data per core	2× 512KB	HD 6250	80:8:4	276	N/A	44.1	1066	5.9	XMZ01AFVB22GV

C-30	Jan 4, 2011			1 (1)	1.2			512KB							9	CMC30AFPB12GT
C-50	Jan 4, 2011			2 (2)	1.0			2× 512KB								CMC50AFPB22GT
C-60	Aug 22, 2011		C0	2 (2)	1.0	1.33		2× 512KB	HD 6290			400				CMC60AFPB22GV

E-240	Jan 4, 2011		B0	1 (1)	1.5	N/A		512KB	HD 6310		500	N/A	80	1066	18	EME240GBB12GT
E-300	Aug 22, 2011			2 (2)	1.3			2× 512KB			488		78			EME300GBB22GV
E-350	Jan 4, 2011				1.6						492		78.7			EME350GBB22GT
E-450	Aug 22, 2011		B0 C0		1.65				HD 6320		508	600	81.2	1333		EME450GBB22GV

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Brazos 2.0: "Ontario", "Zacate" (2012)[edit]

Fabrication 40 nm by TSMC
Socket FT1 (BGA-413)
Based on the Bobcat microarchitecture^[175]
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, ABM, NX bit, AMD64, AMD-V
PowerNow!
DirectX 11 integrated graphics
C-series denote Ontario; and the E-series denotes Zacate
2.50 GT/s UMI (PCIe 1.0 ×4)

Model	Released	Fab	Step.	CPU						GPU					DDR3 Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2	L3	Model	Config	Clock (MHz)	Turbo (MHz)	GFLOPS^[b]
C-70	Sep 15, 2012	40nm	C0	2 (2)	1.0	1.33	32 KB inst. 32 KB data per core	2× 512KB	N/A	HD 7290	80:8:4	276	400	44.1	1066	9	CMC70AFPB22GV

E1-1200	Jun 6, 2012		C0		1.4	N/A				HD 7310		500	N/A	80	1066	18	EM1200GBB22GV
E1-1500	Jan 7, 2013				1.48					HD 7310		529	N/A	84.6	1066
E2-1800	Jun 6, 2012				1.7					HD 7340		523	680	83.6	1333		EM1800GBB22GV
E2-2000	Jan 7, 2013				1.75					HD 7340		538	700	86	1333

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Brazos-T: "Hondo" (2012)[edit]

Fabrication 40 nm by TSMC
Socket FT1 (BGA-413)
Based on the Bobcat microarchitecture^[175]
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
Found in tablet computers
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, ABM, NX bit, AMD64, AMD-V
PowerNow!
DirectX 11 integrated graphics
2.50 GT/s UMI (PCIe 1.0 ×4)

Model	Released	Fab	Step.	CPU				GPU				DDR3 Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
				Cores (threads)	Clock (GHz)	L1	L2	Model	Config	Clock (MHz)	GFLOPS^[b]
Z-60	Oct 9, 2012	40nm	C0	2 (2)	1.0	32KB inst. 32KB data per core	2× 512 KB	HD 6250	80:8:4	276	44.1	1066	4.5	XMZ60AFVB22GV

^ AMD in its technical documentation uses KB, which it defines as "kilobyte" and as equal to 1024 B (i.e., 1 KiB), and MB, which it defines as "megabyte" and as equal to 1024 KB (1 MiB).^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Kabini", "Temash" (2013)[edit]

Fabrication 28 nm by TSMC
Socket FT3 (BGA)
2 to 4 CPU Cores (Jaguar (microarchitecture))
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
Turbo Dock Technology, C6 and CC6 low power states
GPU based on Graphics Core Next (GCN)
AMD Eyefinity multi-monitor for up to two displays

Temash, Elite Mobility APU[edit]

Model	Released	Fab	Step.	CPU					GPU				DDR3L Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Turbo (MHz)
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Turbo (MHz)
A4-1200	May 23, 2013	28nm	KB-A1	2 (2)	1.0	N/A	32 KB inst. 32 KB data per core	1	HD 8180	128:8:4 2 CU	225	N/A	1066	4	AT1200IFJ23HM
A4-1250				2 (2)				1	HD 8210		300		1333	8	AT1250IDJ23HM
A4-1350				4 (4)				2	HD 8210				1066		AT1350IDJ44HM
A6-1450				4 (4)		1.4		2	HD 8250			400	1066		AT1450IDJ44HM

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

Kabini, Mainstream APU[edit]

Model	Released	Fab	Step.	CPU					GPU			DDR3L Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Cache^[a]			Model	Config	Clock (MHz)
				Cores (threads)	Clock (GHz)	L1	L2 (MB)	L3	Model	Config	Clock (MHz)
E1-2100	May 2013	28nm	KB-A1	2 (2)	1.0	32KB inst. 32KB data per core	1	N/A	HD 8210	128:8:4 2 CU	300	1333	9	EM2100ICJ23HM
E1-2200	Feb 2014				1.05				HD 8210		300		9	EM2200ICJ23HM
E1-2500	May 2013				1.4				HD 8240		400		15	EM2500IBJ23HM
E2-3000	May 2013				1.65				HD 8280		450	1600		EM3000IBJ23HM
E2-3800	Feb 2014			4	1.3		2		HD 8280		450			EM3800IBJ44HM
A4-5000	May 2013				1.5				HD 8330		497			AM5000IBJ44HM
A4-5100	Feb 2014				1.55				HD 8330		497			AM5100IBJ44HM
A6-5200	May 2013				2.0				HD 8400		600		25	AM5200IAJ44HM
A4 Pro-3340B	Nov 2014				2.2				HD 8240		400		25	AM334BIAJ44HM

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

"Beema", "Mullins" (2014)[edit]

Fabrication 28 nm by GlobalFoundries
Socket FT3b (BGA)
CPU: 2 to 4 (Puma cores)
- L1 Cache: 32 KB Data per core and 32 KB Instructions per core
GPU based on Graphics Core Next (GCN)
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
Intelligent Turbo Boost
Platform Security Processor, with an integrated ARM Cortex-A5 for TrustZone execution

Mullins, Tablet/2-in-1 APU[edit]

Model	Released	Fab	Step.	CPU						GPU				DDR3L Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config	Clock (MHz)	Turbo (MHz)
				Cores (threads)	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	L3	Model	Config	Clock (MHz)	Turbo (MHz)
E1 Micro-6200T	Q2 2014	28nm	ML-A1	2 (2)	1.0	1.4	32 KB inst. 32 KB data per core	1	N/A	R2	128:8:4 2 CU	300	600	1066	3.95	EM620TIWJ23JB
A4 Micro-6400T				4 (4)	1.0	1.6		2		R3		350	686	1333	4.5	AM640TIVJ44JB
A10 Micro-6700T				4 (4)	1.2	2.2		2		R6		500	N/A	1333	4.5	AM670TIVJ44JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

Beema, Notebook APU[edit]

Model	Released	Fab	Step.	CPU						GPU				DDR3 Memory support	TDP (W)	Part number
				Cores (threads) [FPUs]	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config	Clock (MHz)	Turbo (MHz)
				Cores (threads) [FPUs]	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	L3	Model	Config	Clock (MHz)	Turbo (MHz)
E1-6010	Q2 2014	28nm	ML-A1	2 (2)	1.35	N/A	32 KB inst. 32 KB data per core	1	N/A	R2	128:8:4 2 CU	300	600	(L)1333	10	EM6010IUJ23JB
E1-6015^[176]	Q2 2015			2 (2)	1.4			1						(L)1333	10
E2-6110	Q2 2014			4 (4)	1.5			2						(L)1600	15	EM6110ITJ44JB
A4-6210					1.8					R3		350	686		15	AM6210ITJ44JB
A4-6250J^[177]					2.0					R3		350	686		25
A6-6310					1.8	2.4				R4		300	800	(L)1866	15	AM6310ITJ44JB
A8-6410					2.0					R5				(L)1866		AM6410ITJ44JB
A4 Pro-3350B	May 2016				2.0					R4				1600		AM335BITJ44JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

"Carrizo-L" (2015)[edit]

Fabrication 28 nm by GlobalFoundries
Socket FT3b (BGA), FP4 (µBGA)^[178]
CPU: 2 to 4 (Puma+ cores)
- L1 Cache: 32 KB Data per core and 32 KB Instructions per core
GPU based on Graphics Core Next (GCN)
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
Intelligent Turbo Boost
Platform Security Processor, with an integrated ARM Cortex-A5 for TrustZone execution
All models except A8-7410 available in both laptop and all-in-one desktop versions

Model	Released	Fab	Step.	CPU					GPU				DDR3L Memory support	TDP (W)	Part number
				Cores (threads) [FPUs]	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock	Turbo (MHz)
				Cores (threads) [FPUs]	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock	Turbo (MHz)
E1-7010	May 2015	28nm	ML-A1	2	1.5	N/A	32 KB inst. 32 KB data per core	1	R2	128:8:4 2 CU		400	1333	10	EM7010IUJ23JB EM7010JCY23JB EM7010JCY23JBD
E2-7110				4	1.8	N/A		2	R2			600	1600	12– 25	EM7110ITJ44JB EM7110JBY44JB EM7110JBY44JBD
A4-7210					1.8	2.2			R3			686	1600		AM7210ITJ44JB AM7210JBY44JBD
A6-7310					2.0	2.4			R4			800	1866		AM7310ITJ44JB AM7310JBY44JB AM7310JBY44JBD
A8-7410					2.2	2.5			R5			847	1866		AM7410JBY44JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

"Stoney Ridge" (2016)[edit]

Fabrication 28 nm by GlobalFoundries
Socket FP4^[105] / FT4
2 "Excavator+" x86 CPU cores
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND, Turbo Core
GPU based on Graphics Core Next 3rd Generation with VP9 decoding

Model number	Released	Fab	CPU					GPU				DDR4 Memory support	TDP (W)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config	Clock (MHz)	GFLOPS^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	GFLOPS^[b]
E2-9000e	November 2016	28nm	[1]2	1.5	2.0	96 KB inst. per module 32 KB data per core	1	R2	128:8:4 2 CU	600	153.6	1866	6	EM900EANN23AC
E2-9000	June 2016			1.8	2.2								10	EM9000AKN23AC
E2-9010	June 2016			2.0	2.2								10–15	EM9010AVY23AC
A4-9120	Q2 2017			2.2	2.5			R3		655	167.6	2133	10–15	AM9120AYN23AC
A4-9125	Q2 2018			2.3	2.6			R3		686	175.6	2133	10–15	AM9125AYN23AC
A4-9120C	January 6, 2019			1.6	2.4			R4	192:12:8 3 CU	600	230.4	1866	6	AM912CANN23AC
A6-9200e	November 2016			1.8	2.7							2133	6	AM920EANN23AC
A6-9200	November 2016			2.0	2.8								10	AM9200AKN23AC
A6-9210	June 2016			2.4	2.8								10–15	AM9210AVY23AC
A6-9220	Q2 2017			2.5	2.9					655	251.5		10–15	AM9220AYN23AC
A6-9225	Q2 2018			2.6	3.0					686	263.4		10–15	AM9225AYN23AC
A6-9220C	January 6, 2019			1.8	2.7			R5		720	276.4	1866	6	AM922CANN23AC
A9-9400	November 2016			2.4	3.2					800	307.2	2133	10	AM9400AKN23AC
A9-9410	June 2016			2.9	3.5					800	307.2		10–25	AM9410AFY23AC
A9-9420	Q2 2017			3.0	3.6					847	325.2			AM9420AYN23AC
A9-9425	Q2 2018			3.1	3.7					900	345.6			AM9425AYN23AC
A9-9430^[179]	Q2 2017			3.2	3.5					847	325.2	2400	25	AD9430AJN23AC

Pro A4-4350B	Q1 2018			2.5	2.9					655	251.5	2133	15
Pro A4-5350B	Q1 2020			3.0	3.6					847	325.2
Pro A6-7350B	Q1 2018			3.0	3.6					847	325.2
Pro A6-8350B	Q1 2020			3.1	3.7					900	345.6

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Dalí", "Pollock" (2020)[edit]

Fabrication 14 nm by GlobalFoundries
Socket FP5 / FT5
Two Zen CPU cores
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core

Model	Release date	Fab	CPU						GPU				Socket	PCIe lanes	Memory support	TDP	Part number
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[a]	Clock	Processing power (GFLOPS)^[b]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[a]	Clock	Processing power (GFLOPS)^[b]
AMD 3015e^[180]	July 6, 2020	14 nm	2 (4)	1.2	2.3	64 KB inst. 32 KB data per core	512 KB per core	4 MB	AMD Radeon^(TM) Graphics (Vega)	3 CU	600 MHz		FT5	12 (8+4)	DDR4-1600 single-channel	6 W	AM3015BRP2OFJ
AMD 3020e^[181]	January 6, 2020		2 (2)	1.2	2.6					192:12:4 3 CU	1000 MHz	384	FP5		DDR4-2400 dual-channel		YM3020C7T2OFG
Athlon Silver 3050e^[182]	January 6, 2020		2 (4)	1.4	2.8					192:12:4 3 CU	1000 MHz	384					YM3050C7T2OFG
Athlon PRO 3045B	Q1 2021		2 (2)	2.3	3.2					128:8:4 2 CU	1100 MHz	281.6				12-25 W	YM3045C4T2OFG
Athlon Silver 3050U^[183]	January 6, 2020																YM3050C4T2OFG
Athlon Silver 3050C^[184]	September 22, 2020																YM305CC4T2OFG
Athlon PRO 3145B	Q1 2021		2 (4)	2.6	3.3					192:12:4 3 CU	1000 MHz	384					YM3145C4T2OFG
Athlon Gold 3150U^[185]	January 6, 2020																YM3150C4T2OFG
Athlon Gold 3150C^[186]	September 22, 2020																YM315CC4T2OFG
Ryzen 3 3250U^[187]	January 6, 2020			2.6	3.5						1200 MHz	460.8					YM3250C4T2OFG
Ryzen 3 3250C^[188]	September 22, 2020			2.6	3.5						1200 MHz	460.8					YM325CC4T2OFG

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Embedded APUs[edit]

G-Series[edit]

Brazos: "Ontario" and "Zacate" (2011)[edit]

Fabrication 40 nm
Socket FT1 (BGA-413)
CPU microarchitecture: Bobcat^[189]
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, ABM, NX bit, AMD64, AMD-V
GPU microarchitecture: TeraScale 2 (VLIW5) "Evergreen"
Memory support: single-channel, support up to two DIMMs of DDR3-1333 or DDR3L-1066
5 GT/s UMI

Model	Released	Fab	Step.	CPU				GPU				DDR3 Memory support	TDP (W)	Part number
				Cores (threads)	Clock (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads)	Clock (GHz)	L1	L2	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
G-Series T24L	Mar 1, 2011 May 23, 2011	40nm	B0	1 (1)	0.8 1.0	32 KB inst. 32 KB data per core	512 KB	N/A				1066	5	GET24LFPB12GTE GET24LFQB12GVE
G-Series T30L	Mar 1, 2011 May 23, 2011			1 (1)	1.4		512 KB						18	GET30LGBB12GTE GET30LGBB12GVE
G-Series T48L	Mar 1, 2011 May 23, 2011			2 (2)	1.4		2 × 512 KB						18	GET48LGBB22GTE GET48LGBB22GVE

G-Series T16R	Jun 25, 2012		B0	1 (1)	0.615		512 KB	HD 6250	80:8:4	276	44.1	(L)1066	4.5	GET16RFWB12GVE
G-Series T40R	May 23, 2011			1 (1)	1.0		512 KB			280	44.8	1066	5.5	GET40RFQB12GVE
G-Series T40E	May 23, 2011			2 (2)			2 × 512 KB						6.4	GET40EFQB22GVE
G-Series T40N	Jan 19, 2011 May 23, 2011			2 (2)			2 × 512 KB	HD 6250 HD 6290					9	GET40NFPB22GTE GET40NFPB22GVE
G-Series T40R	May 23, 2011			1 (1)			512 KB	HD 6250					5.5	GET40RFSB12GVE
G-Series T44R	Jan 19, 2011 May 23, 2011			1 (1)	1.2		512 KB						9	GET44RFPB12GTE GET44RFPB12GVE
G-Series T48E	Jun 25, 2012			2 (2)	1.4		2 × 512 KB						18	GET48EGBB22GVE
G-Series T48N	Jan 19, 2011 May 23, 2011			2 (2)	1.4		2 × 512 KB	HD 6310		500 520	80 83.2			GET48NGBB22GTE GET48NGBB22GVE
G-Series T52R	Jan 19, 2011 May 23, 2011			1 (1)	1.5		512 KB	HD 6310		500	80	1066 1333		GET52RGBB12GTE GET52RGBB12GVE
G-Series T56E	Jun 25, 2012			2 (2)	1.65		2 × 512 KB	HD 6250		275	44	1333		GET56EGBB22GVE
G-Series T56N	Jan 19, 2011 May 23, 2011			2 (2)	1.6 1.65		2 × 512 KB	HD 6310 HD 6320		500	80	1066 1333		GET56NGBB22GTE GET56NGBB22GVE

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Kabini" (2013, SoC)[edit]

Fabrication 28 nm
Socket FT3 (769-BGA)^[190]
CPU microarchitecture: Jaguar
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support. No support for FMA (Fused Multiply-Accumulate). Trusted Platform Module (TPM) 1.2 support
GPU microarchitecture: Graphics Core Next (GCN) with Unified Video Decoder 3 (H.264, VC-1, MPEG2, etc.)
Single channel DDR3-1600, 1.25 and 1.35 V voltage level support, support for ECC memory
Integrates Controller Hub functional block, HD audio, 2 SATA channels, USB 2.0 and USB 3.0 (except GX-210JA)

Model	Released	Fab	Step.	CPU				GPU				DDR3 Memory support	TDP (W)	Junction temperature (°C)	Part number
				Cores (threads) [FPUs]	Clock (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads) [FPUs]	Clock (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
GX-210UA	Unknown	28nm	B0	2 (2) [1]	1.0	32 KB inst. 32 KB data per core	1	N/A				1333	8.5	0-90	GE210UIGJ23HM
GX-210JA	Jul 30, 2013							HD 8180E	128:8:4 2 CU	225	57.6	1066	6	0-90	GE210JIHJ23HM
GX-209HA	Unknown							HD 8400E		600	153.6	1066	9	-40-105	GE209HISJ23HM
GX-210HA	Jun 1, 2013							HD 8210E		300	76.8	1333	9	0-90	GE210HICJ23HM
GX-217GA	Jun 1, 2013				1.65			HD 8280E		450	115.2	1600	15	0-90	GE217GIBJ23HM
GX-411GA	Unknown			4 (4) [2]	1.1		2	HD 8210E		300	76.8	1066		-40-105	GE411GIRJ44HM
GX-415GA	Jun 1, 2013				1.5			HD 8330E		500	128	1600		0-90	GE415GIBJ44HM
GX-416RA					1.6			N/A							GE416RIBJ44HM
GX-420CA					2.0			HD 8400E	128:8:4 2 CU	600	153.6		25		GE420CIAJ44HM

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Steppe Eagle" (2014, SoC)[edit]

Fabrication 28 nm
Socket FT3b (769-BGA)
CPU microarchitecture: Puma
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support

Model	Released	Fab	Step.	CPU				GPU				DDR3 Memory support	TDP (W)	Junction temperature (°C)	Part number
				Cores (threads) [FPUs]	Clock (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads) [FPUs]	Clock (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
GX-210JC	June 4, 2014	28nm	ML-A1	2 (2) [1]	1.0	32 KB inst. 32 KB data per core	1	R1E	128:8:4 2 CU	267	68.3	1600	6	-40-105	GE210JIZJ23JB
GX-212JC					1.2			R2E		300	76.8	1333	6	0-90	GE212JIYJ23JB
GX-216HC					1.6			R4E		300	76.8	1066	10	-40-105	GE216HHBJ23JB
GX-222GC					2.2			R5E		655	167.6	1600	15	0-90	GE222GITJ23JB
GX-412HC				4 (4) [2]	1.2		2	R3E		300	76.8	1333	7		GE412HIYJ44JB
GX-424CC				4 (4) [2]	2.4		2	R5E		497	127.2	1866	25		GE424CIXJ44JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Crowned Eagle" (2014, SoC)[edit]

Fabrication 28 nm
Socket FT3b (769-BGA)
CPU microarchitecture: Puma
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
no GPU

Model	Released	Fab	CPU				GPU	DDR3 Memory support	TDP (W)	Junction temperature (°C)	Part number
			Cores (threads) [FPUs]	Clock (GHz)	Cache^[a]
			Cores (threads) [FPUs]	Clock (GHz)	L1	L2 (MB)
GX-224PC	June 4, 2014	28nm	2 (2) [1]	2.4	32 KB inst. 32 KB data per core	1	N/A	1866	25	0-90	GE224PIXJ23JB
GX-410VC			4 (4) [2]	1.0		2		1066	7	-40-105	GE410VIZJ44JB
GX-412TC				1.2				1600	6	0-90	GE412TIYJ44JB
GX-420MC				2.0				1600	17.5	0-90	GE420MIXJ44JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

LX-Family (2016, SoC)[edit]

Fabrication 28 nm
Socket FT3b (769-BGA)
2 Puma x86 cores with 1MB shared L2 cache
L1 Cache: 32 KB Data per core and 32 KB Instructions per core
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AVX, F16C, CLMUL, AES, MOVBE (Move Big-Endian instruction), XSAVE/XSAVEOPT, ABM, BMI1, AMD-V support
GPU microarchitecture: Graphics Core Next (GCN) (1CU) with support for DirectX 11.2
Single channel 64-bit DDR3 memory with ECC
Integrated Controller Hub supports: PCIe® 2.0 4×1, 2 USB3 + 4 USB2 ports, 2 SATA 2.0/3.0 ports

Model	Released	Fab	Step.	CPU				GPU				DDR3 Memory support	TDP (W)	Part number
				Cores (threads) [FPUs]	Clock (GHz)	Cache^[a]		Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
				Cores (threads) [FPUs]	Clock (GHz)	L1	L2 (MB)	Model	Config	Clock (MHz)	Processing power (GFLOPS)^[b]
GX-208JL	Feb 23, 2016	28nm	ML-A1	2	0.8	32 KB inst. 32 KB data per core	1	R1E	64:4:1 1 CU	267	34.1	1333	6	GE208JIVJ23JB
GX-210HL	2017				1.0							1066	7	GE208HIZJ23JB
GX-210JL	Feb 23, 2016											1333	6	GE210JIVJ23JB
GX-210KL	2017											1333	4.5	GE210KIVJ23JB
GX-215GL	Feb 23, 2016				1.5					497	63.6	1600	15	GE215GITJ23JB
GX-218GL	Feb 23, 2016				1.8					497	63.6	1600	15	GE218GITJ23JB

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

I-Family: "Brown Falcon" (2016, SoC)[edit]

Fabrication 28 nm
Socket FP4^[191]
2 or 4 Excavator x86 cores with 1MB shared L2 cache
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND
GPU microarchitecture: Graphics Core Next (GCN) (up to 4 CUs) with support for DirectX 12
Dual channel 64-bit DDR4 or DDR3 memory with ECC
4K × 2K H.265 decode capability and multi format encode and decode
Integrated Controller Hub supports: PCIe 3.0 1×4, PCIe 2/3 4×1, 2 USB3 + 2 USB2 ports, 2 SATA 2.0/3.0 ports

Model	Released	Fab	CPU					GPU				Memory support	TDP (W)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Processing power (GFLOPS)^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Processing power (GFLOPS)^[b]
GX-217GI	Feb 23, 2016	28nm	[1] 2	1.7	2.0	96 KB inst. per module 32 KB data per core	1	R6E	256:16:4 4 CU	758	388	DDR3/DDR4-1600	15	GE217GAAY23KA
GX-420GI^[192]^[193]	2016	28nm	[2] 4	2.0	2.2	96 KB inst. per module 32 KB data per core	2	R6E R7E	256:16:4 4 CU 384:24:4 6 CU	758 626	388 480.7	DDR4-1866	16.1	GE420GAAY43KA

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

J-Family: "Prairie Falcon" (2016, SoC)[edit]

Fabrication 28 nm
Socket FP4^[194]
2 "Excavator+" x86 cores with 1MB shared L2 cache
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND
GPU microarchitecture: Radeon R5E Graphics Core Next (GCN) (up to 3 CUs) with support for DirectX 12
Single channel 64-bit DDR4 or DDR3 memory
4K × 2K H.265 decode capability with 10-bit compatibility and multi format encode and decode
Integrated Controller Hub supports: PCIe 3.0 1×4, PCIe 2/3 4×1, 2 USB3 + 2 USB2 ports, 2 SATA 2.0/3.0 ports

Model	Released	Fab	CPU					GPU					Memory support	TDP (W)	Junction temperature (°C)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Turbo	Processing power (GFLOPS)^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Turbo	Processing power (GFLOPS)^[b]
GX-212JJ	2018	28nm	[1] 2	1.2	1.6	96 KB inst. per module 32 KB data per core	1	R1E	64:4:1 1 CU	600	N/A	76.8	DDR3-1333 DDR4-1600	6– 10	0-90	GE212JAWY23AC
GX-215JJ	2017			1.5	2.0			R2E	128:8:2 2 CU			153.6	DDR3-1600 DDR4-1866	6– 10		GE215JAWY23AC
GX-220IJ	2018			2.0	2.2			R2E	128:8:2 2 CU			153.6	DDR3-1600 DDR4-1866	10– 15		GE220IAVY23AC
GX-224IJ	2017			2.4	2.8			R4E	192:12:3 3 CU			230.4	DDR3-1866 DDR4-2133	10– 15		GE224IAVY23AC

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

R-Series[edit]

Comal: "Trinity" (2012)[edit]

Fabrication 32 nm
Socket FP2 (BGA-827), FS1r2
CPU microarchitecture: Piledriver
L1 Cache: 16 KB Data per core and 64 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AMD64, AMD-V, AES, CLMUL, AVX 1.1, XOP, FMA3, FMA4, F16C,^[29] ABM, BMI1, TBM
GPU microarchitecture: TeraScale 3 (VLIW4) "Northern Islands"
Memory support: dual-channel 1.35 V DDR3L-1600 memory, in addition to regular 1.5 V DDR3
2.5 GT/s UMI
Die size: 246 mm²; Transistors: 1.303 billion
OpenCL 1.1 and OpenGL 4.2 support

Model	Released	Fab	Step.	CPU					GPU					DDR3 Memory support	TDP (W)	Part number
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Turbo (MHz)	Processing power (GFLOPS)^[b]
				[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Turbo (MHz)	Processing power (GFLOPS)^[b]
R-252F	May 21, 2012	32nm	B0	[1] 2	1.9	2.4	64 KB inst. per module 16 KB data per core	1	HD 7400G	192:12:4 3 CU	333	417	127.8	1333	17	RE252FSHE23HJE
R-260H					2.1	2.6		2?	HD 7500G	256:16:8 4 CU	327	424	167.4	1333	17	RE260HSHE24HJE
R-268D					2.5	3.0		1	HD 7420G	192:12:4 3 CU	470	640	180.4	1600	35	RE268DDEC23HJE
R-272F					2.7	3.2		1	HD 7520G	192:12:4 3 CU	497	686	190.8		35	RE272FDEC23HJE
R-452L				[2] 4	1.6	2.4		2 × 2 MB	HD 7600G	256:16:8 4 CU	327	424	167.4		19	RE452LSHE44HJE
R-460H					1.9	2.8			HD 7640G	256:16:8 4 CU	497	655	254.4		35	RE460HDEC44HJE
R-460L					2.0	2.8			HD 7620G	384:24:8 6 CU	360	497	276.4	1333	25	RE460LSIE44HJE
R-464L					2.3	3.2			HD 7660G	384:24:8 6 CU	497	686	381.6	1600	35	RE464LDEC44HJE

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Bald Eagle" (2014)[edit]

Fabrication 28 nm
Socket FP3
Up to 4 Steamroller x86 cores^[195]
L1 Cache: 16 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4a, SSE4.1, SSE4.2, AMD64, AMD-V, AES, CLMUL, AVX 1.1, XOP, FMA3, FMA4, F16C,^[29] ABM, BMI1, TBM
GPU microarchitecture: Graphics Core Next (GCN) (up to 8 CUs) with support for DirectX 11.1 and OpenGL 4.2
Dual channel DDR3 memory with ECC
Unified Video Decode (UVD) 4.2 and Video Coding Engine (VCE) 2.0

Model	Released	Fab	CPU					GPU					DDR3 Memory support	TDP (W)	Junction temperature (°C)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]		Model	Config²	Clock (MHz)	Turbo (MHz)	Processing power (GFLOPS)^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	Model	Config²	Clock (MHz)	Turbo (MHz)	Processing power (GFLOPS)^[b]
RX-219NB	May 20, 2014	28nm	[1] 2	2.2	3.0	96 KB inst. per module 16 KB data per core	1	N/A					1600	15- 17	0-100	RE219NECH23JA
RX-225FB			[1] 2	2.2	3.0		1	R4	192:12:4 3 CU	464	533	178.1	1600	15- 17		RE225FECH23JA
RX-425BB			[2] 4	2.5	3.4		4	R6	384:24:8 6 CU	576	654	442.3	1866	30- 35		RE425BDGH44JA
RX-427BB				2.7	3.6			R7	512:32:8 8 CU	600	686	614.4	2133	30- 35		RE427BDGH44JA
RX-427NB				2.7	3.6			N/A					2133	30- 35		RE427NDGH44JA

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

"Merlin Falcon" (2015, SoC)[edit]

Fabrication 28 nm
Socket FP4
Up to 4 Excavator x86 cores^[196]
L1 Cache: 32 KB Data per core and 96 KB Instructions per module
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, XOP, FMA3, FMA4, F16C, ABM, BMI1, BMI2, TBM, RDRAND
GPU microarchitecture: Graphics Core Next (GCN) (up to 8 CUs) with support for DirectX 12
Dual channel 64-bit DDR4 or DDR3 memory with ECC
Unified Video Decode (UVD) 6 (4K H.265 and H.264 decode) and Video Coding Engine (VCE) 3.1 (4K H.264 encode)
Dedicated AMD Secure Processor supports secure boot with AMD Hardware Validated Boot (HVB)
Integrated FCH featuring PCIe 3.0 USB3.0, SATA3, SD, GPIO, SPI, I2S, I2C, UART

Model	Released	Fab	CPU						GPU					Memory support	TDP (W)	Junction temperature (°C)	Part number
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	Cache^[a]			Model	Config²	Clock (GHz)	Turbo	Processing power (GFLOPS)^[b]
			[Modules/FPUs] Cores/threads	Clock (GHz)	Turbo (GHz)	L1	L2 (MB)	L3	Model	Config²	Clock (GHz)	Turbo	Processing power (GFLOPS)^[b]
RX-216TD	October 21, 2015	28nm	[1] 2	1.6	3.0	96 KB inst. per module 32 KB data per core	1	N/A	N/A					DDR3/DDR4-1600	12- 15	0-90	RE216TAAY23KA
RX-216GD	October 21, 2015		[1] 2		3.0		1		R5	256:?:? 4 CU	0.8	N/A	409.6		12- 15	0-90	RE216GAAY23KA
RX-416GD			[2] 4		2.4		2		R6	384:?:? 6 CU	0.72		552.9		15	-40-105	RE416GATY43KA
RX-418GD	October 21, 2015			1.8	3.2				R6	384:?:? 6 CU	0.8		614.4	DDR3-2133 DDR4-2400	12- 35	0-90	RE418GAAY43KA
RX-421BD				2.1	3.4				R7	512:?:? 8 CU	0.8		819.2				RE421BAAY43KA
RX-421ND				2.1	3.4				N/A								RE421NAAY43KA

^ AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

1000-Series[edit]

V1000-Family: "Great Horned Owl" (2018, SoC)[edit]

Fabrication 14 nm by GlobalFoundries
Up to 4 Zen cores
Socket FP5
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Dual channel DDR4 memory with ECC
Fifth generation GCN based GPU

Model	Release date	Fab	CPU						GPU				DDR4 Memorysupport	Ethernet	TDP	Junctiontemperature (°C)
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
V1500B^[197]	December 2018	GloFo 14LP	4 (8)	2.2	N/A	64 KB inst. 32 KB data per core	512 KB per core	4 MB	N/A				2400 dual-channel	2× 10GbE	12–25 W	0–105
V1780B^[197]	December 2018		4 (8)	3.35	3.6				N/A				3200 dual-channel		35–54 W
V1202B^[197]	February 2018		2 (4)	2.3	3.2				RX Vega 3	192:12:16 3 CU	1.0	384	2400 dual-channel		12–25 W
V1404I^[197]	December 2018		4 (8)	2.0	3.6				RX Vega 8	512:32:16 8 CU	1.1	1126.4				−40 – 105
V1605B^[197]	February 2018			2.0							1.1	1126.4				0–105
V1756B^[197]				3.25							1.3	1331.2	3200 dual-channel		35–54 W
V1807B^[197]				3.35	3.8				RX Vega 11	704:44:16 11 CU	1.3	1830.4	3200 dual-channel		35–54 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

R1000-Family: "Banded Kestrel" (2019, SoC)[edit]

Fabrication 14 nm by GlobalFoundries
Up to 2 Zen cores
Socket FP5
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4a, AMD64, AMD-V, AES, CLMUL, AVX, AVX 1.1, AVX2, FMA3, F16C, ABM, BMI1, BMI2, RDRAND, Turbo Core
Dual channel DDR4 memory with ECC
Fifth generation GCN based GPU

Model	Release date	Fab	CPU							GPU				Memorysupport	TDP
			Cores (threads)	Clock rate (GHz)			Cache			Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	XFR	L1	L2	L3	Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
R1102G ^[198]	February 25, 2020	GloFo 14LP	2 (2)	1.2	2.6	Unknown	64 KB inst. 32 KB data per core	512 KB per core	4 MB	RX Vega 3	192:12:4 3 CU	1.0	384	DDR4-2400 single-channel	6 W
R1305G^[198]	February 25, 2020		2 (4)	1.5	2.8	Unknown								DDR4-2400 dual-channel	8-10 W
R1505G^[198]	April 16, 2019			2.4	3.3	Unknown									12–25 W
R1606G^[198]	April 16, 2019			2.6	3.5	Unknown						1.2	460.8		12–25 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

2000-Series[edit]

V2000-Family: "Grey Hawk" (2020, SoC)[edit]

Fabrication 7 nm by TSMC
Up to 8 Zen 2 cores
Fifth generation GCN based GPU

Model	Release date	Fab	CPU						GPU				Socket	Memorysupport	TDP
			Cores (threads)	Clock rate (GHz)		Cache			Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
			Cores (threads)	Base	Boost	L1	L2	L3	Model	Config^[i]	Clock (GHz)	Processing power (GFLOPS)^[ii]
V2516^[199]^[200]	November 10, 2020^[201]	TSMC 7FF	6 (12)	2.1	3.95	32 KB inst. 32 KB data per core	512 KB per core	8 MB	Radeon Vega 6	384:24:8 6 CU	1.5	1152	FP6	DDR4-3200 dual-channel LPDDR4X-4266 quad-channel	10-25 W
V2546^[199]^[200]			6 (12)	3.0	3.95				Radeon Vega 6	384:24:8 6 CU	1.5	1152			35-54 W
V2718^[199]^[200]			8 (16)	1.7	4.15				Radeon Vega 7	448:28:8 7 CU	1.6	1433.6			10-25 W
V2748^[199]^[200]			8 (16)	2.9	4.25				Radeon Vega 7	448:28:8 7 CU	1.6	1433.6			35-54 W

^ Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

Custom APUs[edit]

As of May 1, 2013, AMD opened the doors of their "semi-custom" business unit.^[202] Since these chips are custom-made for specific customer needs, they vary widely from both consumer-grade APUs and even the other custom-built ones. Some notable examples of semi-custom chips that have come from this sector include the chips from the PlayStation 4 and Xbox One.^[203] So far the size of the integrated GPU in these semi-custom APUs exceed by far the GPU size in the consumer-grade APUs.

Chip (device)	Release date	Fab	Die area (mm2)	CPU				GPU							Memory			Storage	API support	Special features
Chip (device)	Release date	Fab	Die area (mm2)	Archi- tecture	Cores	Clock (GHz)	L2 cache	Archi- tecture	Core config^[a]	Clock (MHz)	GFLOPS ^[b]	Pixel fillrate (GP/s) ^[c]	Texture fillrate (GT/s) ^[d]	Other	Size	Bus type & width	Band- width (GB/s)	Storage	API support	Audio	Other
Liverpool (PS4)	Nov 2013	28 nm	348	Jaguar	8 cores	1.6	2× 2 MiB	GCN 2	1152:72:32 18 CU	800	1843	25.6	57.6	8 ACEs	8 GiB	GDDR5 256-bit	176	3DBD/DVD 1× 2.5" SATA hard drive Easily replaceable hard drive USB 3.0	OpenGL 4.2, GNM, GNMX and PSSL	Dolby Atmos (BD) S/PDIF	PS VR PS4 additional modules HDR10 (except discs)^[e] CEC Optional IR sensor
Durango (Xbox One)	Nov 2013		363			1.75			768:48:16 12 CU	853	1310	13.6	40.9	2 ACEs	32 MiB	ESRAM^[f]	204	3DBD/DVD/CD 1× 2.5" SATA hard drive USB 3.0	Direct3D 11.2 and 12	Fully Dolby Atmos, DTS:X, and Windows Sonic S/PDIF	Xbox One additional modules FreeSync (1) HDMI 1.4 through IR sensor and IR out port Kensington lock
Durango (Xbox One)	Nov 2013		363							853	1310	13.6	40.9	2 ACEs	8 GiB	DDR3 256-bit	68	3DBD/DVD/CD 1× 2.5" SATA hard drive USB 3.0		Fully Dolby Atmos, DTS:X, and Windows Sonic S/PDIF
Edmonton (Xbox One S) ^[204]	Jun 2016	16 nm	240							914	1404	14.6	43.9	2 ACEs	32 MiB	ESRAM	219	4KBD/3DBD/DVD/CD^[g] 1× 2.5" SATA hard drive USB 3.0		Fully Dolby Atmos, DTS:X, and Windows Sonic S/PDIF	Xbox One S additional modules Fully HDR10 Dolby Vision (streaming) FreeSync (1&2) HDMI 1.4 through IR sensor and IR out port Kensington lock
Edmonton (Xbox One S) ^[204]	Jun 2016		240							914	1404	14.6	43.9	2 ACEs	8 GiB	DDR3 256-bit	68	4KBD/3DBD/DVD/CD^[g] 1× 2.5" SATA hard drive USB 3.0		Fully Dolby Atmos, DTS:X, and Windows Sonic S/PDIF
(PS4 Slim)	Sep 2016		208			1.6			1152:72:32 18 CU	800	1843	25.6	57.6	8 ACEs	8 GiB	GDDR5 256-bit	176	3DBD/DVD 1× 2.5" SATA hard drive Easily replaceable hard drive USB 3.0	OpenGL 4.2, GNM, GNMX and PSSL	Dolby Atmos (BD)	PS VR PS4 Slim additional modules HDR10 (except discs) CEC Optional IR sensor
Neo (PS4 Pro) ^[205]^[206]^[207]	Nov 2016		325			2.13		GCN 4Polaris ^[208]	2304:144:32 36 CU	911	4198	58.3	131.2	4 ACEs and 2 HWS Double-rate FP16^[h] checkerboard rendering	8 GiB ^[209]	GDDR5 256-bit	218	3DBD/DVD 1× 2.5" SATA hard drive Easily replaceable hard drive USB 3.0	OpenGL 4.2 (4.5), GNM, GNMX and PSSL	Dolby Atmos (BD) S/PDIF	PS VR PS4 Pro additional modules HDR10 (except discs) Up to 4K@60 Hz CEC Optional IR sensor
Neo (PS4 Pro) ^[205]^[206]^[207]	Nov 2016		325			2.13			2304:144:32 36 CU	911	4198	58.3	131.2		1 GiB	DDR3^[i]	?		OpenGL 4.2 (4.5), GNM, GNMX and PSSL	Dolby Atmos (BD) S/PDIF
Scorpio (Xbox One X) ^[210]^[211]^[212]	Nov 2017		359	CustomizedJaguar		2.3			2560:160:32 40 CU	1172	6001	37.5	187.5	4 ACEs and 2 HWS	12 GiB	GDDR5 384-bit	326	4KBD/3DBD/DVD/CD 1× 2.5" SATA hard drive USB 3.0	Direct3D 11.2 and 12	Fully Dolby Atmos, DTS:X, and Windows Sonic S/PDIF	Xbox One X additional modules Fully HDR10 Dolby Vision (streaming) FreeSync (1&2) Up to 4K@60 Hz HDMI 1.4b through IR sensor and IR out port
Fenghuang (Subor Z+) ^[213]^[214]^[215]	cancelled ^[216]	14 nm ^[217]	397	Zen	4 cores 8 threads	3.0		GCN 5Vega	1536:96:32 24 CU	1300	3994	41.6	124.8	Double-rate FP16	8 GiB	GDDR5 256-bit	154	1× 2.5" SATA SSD 1× 2.5" SATA hard drive Easily replaceable drives USB 3.0	Vulkan 1.1, Direct3D 12.1	S/PDIF	Subor Z Plus additional modules Windows 10 Enterprise LTSC
Oberon (PS5) ^[218]	Nov 2020	7 nm	308	Zen 2	8 cores 16 threads	3.5 (variable)		RDNA 2Big Navi	2304:144:64 36 CU	2233 (variable)	10290 (variable)	142.9	321.6	Double-rate FP16 Real-time ray tracing Primitive shaders Custom 3D audio blocks	16 GiB	GDDR6 256-bit	448	4KBD Custom 5.5 GB/s PCIe 4.0 x4 NVMe SSD PCIe 4.0 M.2 slot Easily replaceable M.2 SSD USB (except PS5 games)	Vulkan 1.2	PS5 TEMPEST 3D AudioTech	PS VR Dedicated DMA controller and I/O coprocessors Custom coherency engines and cache scrubbers Custom decompression block HDR Up to 4K@120 Hz Up to 8K@30 Hz
Anaconda (Xbox Series X)	Nov 2020		360			3.6 (3.8 w/o SMT)			3328:208:64 52 CU	1825	12147	116.8	379.6	Double-rate FP16 Real-time ray tracing Mesh shaders Variable rate shading ANN acceleration	10 GiB	GDDR6 320-bit	560	4KBD Custom 2.4 GB/s NVMe SSD Custom expansion card USB 3.1 (except XSX games)	DirectX 12 Ultimate	Custom spatial audio block MS Project Acoustics Fully Dolby Atmos, DTS:X, and Windows Sonic	Custom decompression block HDR VRR Up to 4K@120 Hz Up to 8K@30 Hz CEC
Anaconda (Xbox Series X)			360			3.6 (3.8 w/o SMT)			3328:208:64 52 CU	1825	12147	116.8	379.6		6 GiB	GDDR6 192-bit^[j]	336
Lockhart (Xbox Series S)			197			3.4 (3.6 w/o SMT)			1280:80:32 20 CU	1565	4006	50.1	125.2		8 GiB	GDDR6 128-bit	224
Lockhart (Xbox Series S)			197			3.4 (3.6 w/o SMT)			1280:80:32 20 CU	1565	4006	50.1	125.2		2 GiB	GDDR6 32-bit	56

^ Unified Shaders : Texture Mapping Units : Render Output Units
^ Precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
^ Pixel fillrate is calculated as the number of ROPs multiplied by the base (or boost) core clock speed.
^ Texture fillrate is calculated as the number of TMUs multiplied by the base (or boost) core clock speed.
^ UHD BD is the only video disc format supporting HDR.
^ Cache
^ "Digital" version does not have an optical drive.
^ Feature preview of Rapid Packed Math, introduced in GCN 5 Vega.
^ Swap
^ A plain 320-bit 20GiB version could be made by just replacing four 1 GiB GDDR6 chips by 2 GiB ones.

Notes[edit]

Note 1: The clock multiplier applies to the 200 MHz base clock. (AMD default clock base is 200 MHz)
Note 2: Unified shader processors (USPs): Texture mapping units (TMUs): Render output units (ROPs). 1 CU (Compute Unit) = 64 USPs: 4 TMUs : 1 ROPs
Note 3: Models enabled by Turbo Core technology, up to 10% clock speed increase is planned. With CPU boost, only one core of a dual-core model has boost enabled.
Note 4: K models feature an unlocked multiplier and overclockable GPU.
Note 5: One AMD module consists of two integer cores and two FPUs, which can be combined into one wide FPU for executing certain instructions, such as FMA. The two cores share certain resources, but are two separate units.

References[edit]

^ "AMD Announces the 7th Generation APU: Excavator mk2 in Bristol Ridge and Stoney Ridge for Notebooks". 31 May 2016. Retrieved 3 January 2020.
^ "AMD Mobile "Carrizo" Family of APUs Designed to Deliver Significant Leap in Performance, Energy Efficiency in 2015" (Press release). 20 November 2014. Retrieved 16 February 2015.
^ "The Mobile CPU Comparison Guide Rev. 13.0 Page 5 : AMD Mobile CPU Full List". TechARP.com. Retrieved 13 December 2017.
^ a b "AMD VEGA10 and VEGA11 GPUs spotted in OpenCL driver". VideoCardz.com. Retrieved 6 June 2017.
^ Cutress, Ian (1 February 2018). "Zen Cores and Vega: Ryzen APUs for AM4 – AMD Tech Day at CES: 2018 Roadmap Revealed, with Ryzen APUs, Zen+ on 12nm, Vega on 7nm". Anandtech. Retrieved 7 February 2018.
^ Larabel, Michael (17 November 2017). "Radeon VCN Encode Support Lands in Mesa 17.4 Git". Phoronix. Retrieved 20 November 2017.
^ Liu, Leo (2020-09-04). "Add Renoir VCN decode support". Retrieved 2020-09-11. It has same VCN2.x block as Navi1x
^ Tony Chen; Jason Greaves, "AMD's Graphics Core Next (GCN) Architecture" (PDF), AMD, retrieved 13 August 2016
^ "A technical look at AMD's Kaveri architecture". Semi Accurate. Retrieved 6 July 2014.
^ "How do I connect three or More Monitors to an AMD Radeon™ HD 5000, HD 6000, and HD 7000 Series Graphics Card?". AMD. Retrieved 8 December 2014.
^ Airlie, David (26 November 2009). "DisplayPort supported by KMS driver mainlined into Linux kernel 2.6.33". Retrieved 16 January 2016.
^ "Radeon feature matrix". freedesktop.org. Retrieved 10 January 2016.
^ Deucher, Alexander (16 September 2015). "XDC2015: AMDGPU" (PDF). Retrieved 16 January 2016.
^ a b Michel Dänzer (17 November 2016). "[ANNOUNCE] xf86-video-amdgpu 1.2.0". lists.x.org.
^ "Conformant Products - The Khronos Group Inc". The Khronos Group. Retrieved 2019-06-06.
^ "Conformant Products - The Khronos Group Inc". The Khronos Group. Retrieved 2019-06-06.
^ "GPU-Tech.org - Catalyst 11.10 WHQL - First official Battlefield 3 driver for Radeon cards". www.gpu-tech.org.
^ "AMD Radeon Software Crimson Edition Beta". AMD. Retrieved 2018-04-20.
^ "Mesamatrix". mesamatrix.net. Retrieved 2018-04-22.
^ "RadeonFeature". X.Org Foundation. Retrieved 2018-04-20.
^ "Graphics Core Next: The Southern Islands Architecture". Tom's Hardware. 2011-12-21. Retrieved 2013-06-26.
^ "AMD Clarifies 2013 Radeon Plans". Tom's Hardware. 2013-02-20. Retrieved 2013-06-26.
^ "Radeon VEGA Frontier Edition". AMD. 2017-05-30. Retrieved 2017-06-30.
^ "AMD launches A-Series and the first 32nm Athlon II X4 CPUs". Retrieved 2013-11-10.
^ Theo Valich (28 May 2012). "AMD Comes Clean on Transistor Numbers With FX, Fusion Processors". Retrieved 23 August 2013.
^ Anand Lal Shimpi (27 September 2012). "AMD A10-5800K & A8-5600K Review: Trinity on the Desktop, Part 1". Retrieved 23 August 2013.
^ a b c d e f g h i j k l m n o p q r s t u v w x y z aa ab ac ad ae af ag ah "Processor Programming Reference (PPR) for AMD Family 17h Model 01h, Revision B1 Processors" (PDF). AMD Technical Documentation. AMD Developer Central: Advanced Micro Devices, Inc. 2017-04-15. p. 25. Retrieved 2019-11-01.
^ "Trinity Improvements Include Updated Piledriver Cores and VLIW4 GPUs". Retrieved 2013-11-10.
^ a b c "AMD detonates Trinity: Behold Bulldozer's second coming - ExtremeTech". Retrieved 2017-10-07.
^ "AMD Trinity On The Desktop: A10, A8, And A6 Get Benchmarked!—Trinity: Coming Soon To A Desktop Near You". Retrieved 2013-11-10.
^ "AMD Trinity for Desktops. Part 1: Graphics Core". X-bit labs. 2012-09-27. Archived from the original on October 11, 2012.
^ "Review: AMD A10-5800K Dual Graphics evaluation—CPU". Retrieved 2013-11-10.
^ "The AMD A8-3850 Review: Llano on the Desktop". Retrieved 2013-11-10.
^ "Product Search Results—Bottom Line Telecommunications". Bottom Line Telecommunications Corporation. Retrieved 2013-11-10.
^ a b "AMD Sempron CPU". Retrieved 2015-03-02.
^ Альберт Шаповалов (10 September 2014). "Обзор и тестирование процессора AMD Athlon X2 340". Ru.gecid.com/ (in Russian). Retrieved 2016-09-12.
^ Hassan Mujtaba. "AMD A10-6800K and A10-6700 "Richland" APU Review". Wccftech. Retrieved 20 March 2020.
^ a b c "AMD Athlon Processors". Retrieved 2015-03-02.
^ btarunr (23 March 2014). "AMD FX-670K CPU Shows Up in the Wild". TechPowerUp.
^ Anton Shilov (2013-05-30). "AMD's Next-Gen "Kaveri" APUs Will Require New Mainboards". Retrieved 2014-12-17.
^ "AMD Godavari core". www.cpu-world.com. Retrieved 2018-09-16.
^ Joel Hruska. "AMD Kaveri A10-7850K and A8-7600 review: Was it worth the wait for the first true heterogeneous chip?". ExtremeTech. Retrieved 20 March 2020.
^ a b Hassan Mujtaba. "AMD Kaveri APU Architecture Detailed". Retrieved 2015-03-15.
^ a b "A technical look at AMD's Kaveri architecture". SemiAccurate. 2014-01-15.
^ a b c "AMD to add ARM processors to boost chip security". June 14, 2012. Retrieved September 3, 2013.
^ a b c "AMD and ARM Fusion redefine beyond x86". Archived from the original on 2013-11-05. Retrieved 2013-11-10.
^ a b "Carrizo presentation, page 12 - Carrizo is the 1st ARM Trustzone capable performance APU" (PDF). Retrieved January 13, 2020.
^ "AMD A10-7850K Graphics Performance". Retrieved April 2, 2014.
^ "AMD A8-7600 Kaveri APU review - The Embedded GPU - HSA & hUMA". 2014-01-14.
^ Gennadiy Shvets (18 October 2014). "HP offers desktop PCs with AMD FX-770K Kaveri processor". OFweek. Archived from the original on 13 May 2018. Retrieved 23 March 2016.
^ "ASRock - FM2+ CPU Support List". asrock.com. Retrieved 18 October 2020.
^ 电脑维修技术网. "AMD APU A8-7500 CPU怎么样？". pc811.com. Retrieved 18 October 2020.
^ Hassan Mujtaba (26 August 2015). "AMD Details Carrizo APUs Energy Efficient Design at Hot Chips 2015 – 28nm Bulk High Density Design With 3.1 Billion Transistors, 250mm2 Die". Wccftech. Retrieved 20 March 2020.
^ "AMD quietly launches new Carrizo APU: A8-7680 processor". 26 October 2018. Retrieved 29 June 2019.
^ Cutress, Ian (28 October 2018). "Day of the Dead: AMD Releases new Carrizo FM2+ APU, the A8-7680". Retrieved 29 June 2019.
^ Cutress, Ian (23 September 2016). "AMD 7th Gen Bristol Ridge and AM4 Analysis". Anandtech.com. Retrieved 23 September 2016.
^ Sang-ho, Lee (19 September 2016). "AMD Final Heavy Equipment X Carrier ZEN Bristol Ridge A12-9800 platform change". BodNara Korea. Retrieved 12 November 2016.
^ "Processor Programming Reference (PPR) for AMD Family 17h Model 01h, Revision B1 Processors" (PDF). AMD Technical Documentation. AMD Developer Central: Advanced Micro Devices, Inc. 2017-04-15. p. 25. Retrieved 2019-11-01.
^ "AMD Athlon™ 200GE Processor with Radeon™ Vega 3 Graphics". AMD.
^ "AMD Athlon™ PRO 200GE APU". AMD.
^ "AMD Athlon™ 220GE Processor with Radeon™ Vega 3 Graphics".
^ "AMD Athlon™ 240GE Processor with Radeon™ Vega 3 Graphics". AMD.
^ "AMD Athlon™ 3000G Processor with Radeon™ Graphics". AMD.
^ "AMD Athlon™ 300GE".
^ "AMD Athlon™ Silver 3050GE".
^ "AMD Ryzen™ 3 2200GE with Radeon™ Vega 8 Graphics". AMD.
^ "AMD Ryzen™ 3 PRO 2200GE Processor with Radeon™ Vega 8 Graphics".
^ "AMD Ryzen™ 3 PRO 2200G Processor with Radeon™ Vega 8 Graphics". www.amd.com.
^ "Specs". www.amd.com. Retrieved 2019-06-10.
^ "Specs". www.amd.com. Retrieved 2019-06-10.
^ "AMD Ryzen™ 5 2400G". Retrieved 2018-01-19.
^ "AMD's 2nd-gen Ryzen is coming in April, desktop Ryzen APUs arrive February 12". TechSpot. Retrieved 2019-06-10.
^ Peter Bright - Jan 8, 2018 9:50 pm UTC (2018-01-08). "AMD's 2018 roadmap: Desktop APUs in February, second-generation Ryzen in April". Ars Technica. Retrieved 2019-06-10.
^ "Specs". www.amd.com. Retrieved 2019-06-10.
^ "AMD Athlon PRO 300GE".
^ "AMD Athlon Silver PRO 3125GE".
^ "AMD Athlon Gold 3150GE".
^ "AMD Athlon Gold PRO 3150GE".
^ "AMD Athlon Gold 3150G".
^ "AMD Athlon Gold PRO 3150G".
^ "AMD Ryzen 3 3200GE".
^ "AMD Ryzen 3 PRO 3200GE".
^ "AMD Ryzen 3 3200G with Radeon Vega 8 Graphics".
^ "AMD Ryzen 3 PRO 3200G".
^ "AMD Ryzen 5 PRO 3350GE".
^ "AMD Ryzen 5 PRO 3350G".
^ "AMD Ryzen 5 3400GE".
^ "AMD Ryzen 5 PRO 3400GE".
^ "AMD Ryzen 5 3400G".
^ "AMD Ryzen 5 3400G".
^ a b c d e f g h i j k l "AMD Ryzen 4000 Series Desktop Processors with AMD Radeon Graphics Set to Deliver Breakthrough Performance for Commercial and Consumer Desktop PCs".
^ "AMD Ryzen 3 5300GE". AMD.
^ a b c "AMD Ryzen 5000G APUs: OEM Only For Now, Full Release Later This Year". AnandTech.
^ "AMD Ryzen 3 5300G". AMD.
^ "AMD Ryzen 5 5600GE". AMD.
^ "AMD Ryzen 5 5600G". AMD.
^ "AMD Ryzen 7 5700GE". AMD.
^ "AMD Ryzen 7 5700G". AMD.
^ Kennedy, Patrick (5 June 2017). "New HPE ProLiant MicroServer Gen10 Powered by AMD Opteron X3000 APUs". Retrieved 5 June 2017.
^ "Opteron". AMD Opteron. AMD. Retrieved 5 June 2017.
^ a b "AMD lists A8-4557M and A10-4657M mobile APUs". www.cpu-world.com. Retrieved 2018-09-17.
^ "AMD intros 35W Richland mobile APUs". Retrieved 2013-11-10.
^ Poeter, Damon. (2013-03-12) AMD Bakes New Interface Capabilities Into Richland APUs | News & Opinion
^ "Archived copy". Archived from the original on 2013-12-03. Retrieved 2013-11-26.CS1 maint: archived copy as title (link)
^ a b Cutress, Ian (1 June 2016). "AMD Announces 7th Generation APU". Anandtech.com. Retrieved 1 June 2016.
^ "AMD A10-9620P SoC - Benchmarks and Specs". Notebookcheck.net. Retrieved 20 July 2018.
^ "AMD A12-9720P SoC - Benchmarks and Specs". Notebookcheck.net. Retrieved 20 July 2018.
^ "HP Pavilion 17 - HP® Official Store". Store.hp.com. Retrieved 20 July 2018.
^ "AMD Athlon™ PRO 200U Mobile Processor with Radeon™ Vega 3 Graphics". AMD.
^ "AMD Athlon™ 300U Mobile Processor with Radeon™ Vega 3 Graphics". AMD.
^ "AMD Ryzen™ 3 2200U". AMD.
^ "AMD Ryzen™ 3 3200U Mobile Processor with Radeon™ Vega 3 Graphics". AMD.
^ "AMD Ryzen™ 3 2300U". AMD.
^ "AMD Ryzen™ 3 PRO 2300U". AMD.
^ "AMD Ryzen™ 5 2500U". AMD.
^ "AMD Ryzen™ 5 PRO 2500U". AMD.
^ "AMD Ryzen™ 5 2600H Mobile Processor with Radeon™ Vega 8 Graphics".
^ "AMD Ryzen™ 7 2700U". AMD.
^ "AMD Ryzen™ 7 PRO 2700U". AMD.
^ "AMD Ryzen™ 7 2800H Mobile Processor with Radeon™ RX Vega 11 Graphics". AMD.
^ "AMD Ryzen™ 3 3300U Mobile Processor with Radeon™ Vega 6 Graphics". Retrieved 2019-01-06.
^ "AMD Radeon Vega 6 Mobile Specs | TechPowerUp GPU Database". Techpowerup.com.
^ "AMD Ryzen™ 3 PRO 3300U Mobile Processor with Radeon™ Vega 6 Graphics".
^ "AMD Ryzen™ 3 3350U". AMD.
^ "AMD Ryzen™ 5 3450U Processor".
^ "AMD Radeon Vega 8 Specs | TechPowerUp GPU Database". Techpowerup.com.
^ "AMD Ryzen™ 5 3500U Mobile Processor with Radeon™ Vega 8 Graphics".
^ "AMD Ryzen™ 5 PRO 3500U Mobile Processor with Radeon™ Vega 8 Graphics".
^ "AMD Ryzen™ 5 3500C".
^ "AMD Ryzen™ 5 3550H Mobile Processor with Radeon™ Vega 8 Graphics". Retrieved 8 January 2018.
^ "AMD Ryzen™ 5 3580U Microsoft Surface® Edition".
^ "AMD Ryzen™ 7 3700U Mobile Processor with Radeon™ RX Vega 10 Graphics".
^ "AMD Radeon RX Vega 10 Mobile Specs | TechPowerUp GPU Database". Techpowerup.com.
^ "AMD Ryzen™ 7 PRO 3700U Mobile Processor with Radeon™ Vega 10 Graphics".
^ "AMD Ryzen™ 7 3700C".
^ "AMD Ryzen™ 7 3750H Mobile Processor with Radeon™ RX Vega 10 Graphics".
^ "AMD Ryzen™ 7 3780U Microsoft Surface® Edition".
^ Cutress, Ian (2020-01-06). "AMD Ryzen 4000 Mobile APUs: 7nm, 8-core on both 15W and 45W, Coming Q1". anandtech.com. AnandTech. Retrieved 2020-01-07.
^ Alcorn, Paul (2020-01-07). "AMD Launches Threadripper 3990X and Ryzen 4000 'Renoir' APUs". tomshardware.com. Tom's Hardware. Retrieved 2020-01-07.
^ Gartenberg, Chaim (2020-01-06). "AMD's 7nm Ryzen 4000 CPUs are here to take on Intel's 10nm Ice Lake laptop chips". theverge.com. The Verge. Retrieved 2020-01-07.
^ a b https://www.techpowerup.com/264801/amd-renoir-die-shot-pictured
^ "AMD Ryzen 3 4300U". AMD.
^ "AMD Ryzen 3 PRO 4450U". AMD.
^ "AMD Ryzen 5 4500U". AMD.
^ "AMD Ryzen 5 4600U". AMD.
^ "AMD Ryzen 5 PRO 4650U". AMD.
^ CoveMiner. "Surface Laptop 4 processors technical overview - Surface". docs.microsoft.com. Retrieved 2021-04-14.
^ "AMD Ryzen 5 4600HS". AMD.
^ "AMD Ryzen 5 4600H". AMD.
^ "AMD Ryzen 7 4700U". AMD.
^ "AMD Ryzen 7 PRO 4750U". AMD.
^ "AMD Ryzen 7 4800U". AMD.
^ CoveMiner. "Surface Laptop 4 processors technical overview - Surface". docs.microsoft.com. Retrieved 2021-04-14.
^ "AMD Ryzen 7 4800HS". AMD.
^ "AMD Ryzen 7 4800H". AMD.
^ "AMD Ryzen 9 4900HS". AMD.
^ "AMD Ryzen 9 4900H". AMD.
^ "AMD Ryzen™ 3 5300U". AMD.
^ "AMD Ryzen™ 5 5500U". AMD.
^ "AMD Ryzen™ 7 5700U". AMD.
^ "AMD Ryzen™ 3 5400U". AMD.
^ "AMD Ryzen™ 3 PRO 5450U". AMD.
^ "AMD Ryzen™ 5 5600U". AMD.
^ "AMD Ryzen™ 5 PRO 5650U". AMD.
^ "AMD Ryzen™ 5 5600H". AMD.
^ "AMD Ryzen™ 5 5600HS". AMD.
^ "AMD Ryzen™ 7 5800U". AMD.
^ "AMD Ryzen™ 7 PRO 5850U". AMD.
^ "AMD Ryzen™ 7 5800H". AMD.
^ "AMD Ryzen™ 7 5800HS". AMD.
^ "AMD Ryzen™ 9 5900HS". AMD.
^ "AMD Ryzen™ 9 5900HX". AMD.
^ "AMD Ryzen™ 9 5980HS". AMD.
^ "AMD Ryzen™ 9 5980HX". AMD.
^ a b c Shimpi, Anand Lal. "Previewing AMD's Brazos, Part 1: More Details on Zacate/Ontario and Fusion". Anandtech.com. Retrieved 20 July 2018.
^ "Archived copy". Archived from the original on 2015-05-27. Retrieved 2015-05-26.CS1 maint: archived copy as title (link)
^ "HP ProDesk 405 G2 Microtower-PC". Retrieved 2015-02-24.
^ Cutress, Ian. "AMD's Carrizo-L APUs Unveiled: 12-25W Quad Core Puma+". Anandtech.com. Retrieved 20 July 2018.
^ "HP Pavilion Desktops - HP® Official Store". Store.hp.com. Retrieved 20 July 2018.
^ "AMD 3015e". AMD.com.
^ "AMD 3020e". AMD.com.
^ "AMD Athlon™ Silver 3050e". AMD.com.
^ "AMD Athlon™ Silver 3050U". Retrieved 2020-01-07.
^ "AMD Athlon™ Silver 3050C".
^ "AMD Athlon™ Gold 3150U". Retrieved 2020-01-08.
^ "AMD Athlon™ Gold 3150C".
^ "AMD Ryzen™ 3 3250U". Retrieved 2020-01-08.
^ "AMD Ryzen™ 3 3250C".
^ "Welcome to AMD - Processors - Graphics and Technology - AMD". Amd.com. Retrieved 20 July 2018.
^ "Embedded Products - High Performance GPU - AMD". Amd.com. Retrieved 20 July 2018.
^ https://www.amd.com/Documents/I-Family-Product-Brief.pdf
^ https://store.hp.com/us/en/pdp/hp-t630-thin-client-p-2zv00at-aba-1?pStoreID=epp
^ http://www.cpu-world.com/CPUs/Bulldozer/AMD-G-Series%20GX-420GI.html
^ https://www.amd.com/Documents/J-Family-Product-Brief.pdf
^ https://www.amd.com/Documents/2nd_Gen_Rseries_Product_Brief.pdf
^ https://www.amd.com/Documents/merlin-falcon-product-brief.pdf
^ a b c d e f g "Embedded Processor Specifications". AMD.
^ a b c d "Embedded Processor Specifications". AMD.
^ a b c d "Embedded Processor Specifications". AMD.
^ a b c d "Product Brief: AMD Ryzen™ Embedded V2000 Processor Family" (PDF).
^ "AMD Unveils AMD Ryzen™ Embedded V2000 Processors with Enhanced Performance and Power Efficiency". AMD.
^ "AMD Establishes Semi-Custom Business Unit to Create Tailored Products with Customer-Specific IP". Archived from the original on 2013-10-01. Retrieved 2013-11-10.
^ "Three for three: How AMD won the war for the heart of next-gen consoles". Retrieved 2013-11-10.
^ MACHKOVECH, SAM (2 August 2016). "Microsoft hid performance boosts for old games in Xbox One S, told no one". Ars Technica. Retrieved 2 August 2016.
^ Walton, Mark (10 August 2016). "PS4 Neo: Sony confirms PlayStation event for September 7". Ars Technica. Retrieved 10 August 2016.
^ Walton, Mark (19 April 2016). "Sony PS4K is codenamed NEO, features upgraded CPU, GPU, RAM—report". Ars Technica. Retrieved 10 August 2016.
^ Smith, Ryan (8 September 2016). "Analyzing Sony's Playstation 4 Pro Hardware Reveal: What Lies Beneath". Anandtech. Retrieved 8 September 2016.
^ Freedman, Andrew (3 November 2017). "Xbox One X vs. PlayStation 4 Pro: Which Powerhouse Should You Get?". Tom's Guide. Retrieved 3 November 2017.
^ "PS4 Pro's additional RAM frees up memory for game developers". Polygon. Retrieved 2018-11-23.
^ http://www.anandtech.com/show/11536/microsofts-project-scorpio-get-a-launch-date-xbox-one-x-499-november-7th
^ https://arstechnica.com/gaming/2017/04/xbox-scorpio-hardware-specs/
^ Cutress, Ian (21 August 2017). "Hot Chips: Microsoft Xbox One X Scoprio Engine Live Blog". Anandtech. Retrieved 21 August 2017.
^ Cutress, Ian (3 August 2018). "AMD Creates Quad Core Zen SoC with 24 Vega CUs for Chinese Consoles". Anandtech.
^ Cutress, Ian (6 August 2018). "More Details About the ZhongShan Subor Z+ Console, with Custom AMD Ryzen SoC". Anandtech.
^ Leadbetter, Richard (15 September 2018). "Hands-on with the Subor Z-Plus: AMD tech tested in new Chinese console". Retrieved 28 October 2018.
^ Judd, Will (16 May 2019). "The Subor Z+ console team has disbanded - but it's not game over yet". Gamer Network.
^ Leadbetter, Leadbetter (15 September 2018). Hands-On: Subor Z Plus Chinese PC/Console Hybrid - Ryzen+Vega AMD Analysis!. Eurogamer. Event occurs at 2 minutes 2 seconds. Retrieved 28 October 2018.
^ Smith, Ryan (16 April 2019). "Sony Teases Next-Gen PlayStation: Custom AMD Chip with Zen 2 CPU & Navi GPU, SSD Too". Anandtech.

External links[edit]

AMD Accelerated Processing Units official website
Technical specification AMD products
AMD products and technologies

[3] For FM2+ Excavator models: A8-7680, A6-7480 & Athlon X4 845.

[nodedef-5] A PC would be one node.

[apudef-6] An APU combines a CPU and a GPU. Both have cores.

[iommubios-7] Requires firmware support.

[sse4a-8] No SSE4. No SSSE3.

[SFLOPS-10] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[11] Unified shaders : texture mapping units : render output units

[DRM-17] To play protected video content, it also requires card, operating system, driver, and application support. A compatible HDCP display is also needed for this. HDCP is mandatory for the output of certain audio formats, placing additional constraints on the multimedia setup.

[19] To feed more than two displays, the additional panels must have native DisplayPort support.^[10] Alternatively active DisplayPort-to-DVI/HDMI/VGA adapters can be employed.

[drm-20] DRM (Direct Rendering Manager) is a component of the Linux kernel. Support in this table refers to the most current version.

[r100_shader-27] Radeon 7000 Series has programmable pixel shaders, but do not fully comply with DirectX 8 or Pixel Shader 1.0. See article on R100's pixel shaders.

[nonpot-28] These series do not fully comply with OpenGL 2+ as the hardware does not support all types of non-power-of-two (NPOT) textures.

[nofp64-33] OpenGL 4+ compliance requires supporting FP64 shaders and these are emulated on some TeraScale chips using 32-bit hardware.

[kib-41] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-42] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[kib-50] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-51] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[kib-55] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-56] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[kib-59] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-60] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[kib-71] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-72] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[box-77] With cooler if available.

[kib-78] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-79] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[cooler-83] A box without cooler might also be available (WOF).

[box-84] With cooler if available.

[kib-85] AMD in its technical documentation uses KB, which it defines as Kilobyte and as equal to 1024 bytes, and MB, which it defines as Megabyte and as equal to 1024 KB.^[27]

[SFLOPS-86] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

[kib-90] AMD defines 1 kilobyte (KB) as 1024 bytes, and 1 megabyte (MB) as 1024 kilobytes.^[58]

[cconfig-91] Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)

[SFLOPS-92] Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.

List of AMD accelerated processing units

Features overview[edit]

Graphics API overview[edit]

Desktop APUs[edit]

Lynx: "Llano" (2011)[edit]

Virgo: "Trinity" (2012)[edit]

"Richland" (2013)[edit]

"Kabini" (2013, SoC)[edit]

"Kaveri" (2014) & "Godavari" (2015)[edit]

"Carrizo" (2016)[edit]

"Bristol Ridge" (2016)[edit]

"Raven Ridge" (2018)[edit]

"Picasso" (2019)[edit]

"Renoir" (2020)[edit]

"Cezanne" (2021)[edit]

Server APUs[edit]

Opteron X2100-series "Kyoto" (2013)[edit]

Opteron X3000-series "Toronto" (2017)[edit]

Mobile APUs[edit]

Sabine: "Llano" (2011)[edit]

Comal: "Trinity" (2012)[edit]

"Richland" (2013)[edit]

"Kaveri" (2014)[edit]

"Carrizo" (2015)[edit]

"Bristol Ridge" (2016)[edit]

"Raven Ridge" (2017)[edit]

"Picasso" (2019)[edit]

"Renoir" (2020)[edit]

"Lucienne" (2021)[edit]

"Cezanne" (2021)[edit]

Ultra-mobile APUs[edit]

Brazos: "Desna", "Ontario", "Zacate" (2011)[edit]

Brazos 2.0: "Ontario", "Zacate" (2012)[edit]

Brazos-T: "Hondo" (2012)[edit]

"Kabini", "Temash" (2013)[edit]

Temash, Elite Mobility APU[edit]

Kabini, Mainstream APU[edit]

"Beema", "Mullins" (2014)[edit]

Mullins, Tablet/2-in-1 APU[edit]

Beema, Notebook APU[edit]

"Carrizo-L" (2015)[edit]

"Stoney Ridge" (2016)[edit]

"Dalí", "Pollock" (2020)[edit]

Embedded APUs[edit]

G-Series[edit]

Brazos: "Ontario" and "Zacate" (2011)[edit]

"Kabini" (2013, SoC)[edit]

"Steppe Eagle" (2014, SoC)[edit]

"Crowned Eagle" (2014, SoC)[edit]

LX-Family (2016, SoC)[edit]

I-Family: "Brown Falcon" (2016, SoC)[edit]

J-Family: "Prairie Falcon" (2016, SoC)[edit]

R-Series[edit]

Comal: "Trinity" (2012)[edit]

"Bald Eagle" (2014)[edit]

"Merlin Falcon" (2015, SoC)[edit]

1000-Series[edit]

V1000-Family: "Great Horned Owl" (2018, SoC)[edit]

R1000-Family: "Banded Kestrel" (2019, SoC)[edit]

2000-Series[edit]

V2000-Family: "Grey Hawk" (2020, SoC)[edit]

Custom APUs[edit]

See also[edit]

Notes[edit]

References[edit]

External links[edit]