neeyik.info
The charts are categorised by the highest DirectX support the model offers - please use the links below to find the appropriate model.  For information about what the specification terms mean, please click here.
	DX6 or earlier	DX7 (TnL)	DX8 (SM1.1)	DX9 (SM2.0)	DX9 (SM3.0)	DX10 (SM4.0)	DX11 (SM5.0)

Model	ATI
	Radeon     5450	Radeon     5550	Radeon     5570	Radeon     5670	Radeon     5750	Radeon     5770	Radeon     5830	Radeon     5850
Core
Codename	Cedar	Redwood	Redwood	Redwood	Juniper	Juniper	Cypress	Cypress
Process scale (nm)	40	40	40	40	40	40	40	40
Transistor count (millions)	292	627	627	627	1040	1040	2150	2150
Core clock speed (MHz)	650	550	650	775	700	850	800	725
Component execution units 1	80	320	400	400	720	800	1120	1440
Vertex shader version (best)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Total effective no. of texture units	8	16	20	20	32	40	56	72
Max pixels output per clock (colour/depth)	4 / 16	8 / 32	8 / 32	8 / 32	16 / 64	16 / 64	16 / 64	32 / 128
Pixel shader version (bset)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Overall DirectX level	11	11	11	11	11	11	11	11
Memory
Interface width (bits)	64	128	128	128	128	128	256	256
Interface type	DDR2 / DDR3	DDR3 / GDDR5	DDR3	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5
Clock speed (MHz)	800	800	900	1000	1150	1200	1000	1000
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	3.25	11.0	16.3	19.4	31.5	42.5	56.0	65.3
Triangle output rate (Gtris/sec)	0.65	0.55	0.65	0.77	0.65	0.85	0.80	0.72
Pixel processing rate (Gpixels/sec)	13.0	44.0	65.0	77.5	126.0	170.0	224.0	261.0
Texel applied rate (Gtexels/sec)	5.2	8.8	13.0	15.5	28.0	34.0	44.8	52.2
Pixel output rate (Gpixels/sec)	2.6	4.4	5.2	6.2	11.2	13.6	12.8	23.2
Memory bandwidth (GB/sec)	6.4	25.6	28.8	64.0	73.6	76.8	128.0	128.0

Model	ATI
	Radeon     5870	Radeon     5970
Core
Codename	Cypress	Hemlock
Process scale (nm)	40	40
Transistor count (millions)	2150	2 x 2150
Core clock speed (MHz)	850	725
Component execution units 1	1600	2 x 1600
Vertex shader version (best)	5.0	5.0
Total effective no. of texture units	80	2 x 80
Max pixels output per clock (colour/depth)	32 / 128	2 x 32 / 128
Pixel shader version (bset)	5.0	5.0
Overall DirectX level	11	11
Memory
Interface width (bits)	256	2 x 256
Interface type	GDDR5	GDDR5
Clock speed (MHz)	1200	1000
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	85.0	145.0 2
Triangle output rate (Gtris/sec)	0.85	1.45 2
Pixel processing rate (Gpixels/sec)	340.0	580.0 2
Texel applied rate (Gtexels/sec)	68.0	116.0 2
Pixel output rate (Gpixels/sec)	27.2	46.4 2
Memory bandwidth (GB/sec)	153.6	256.0 2

Model	AMD
	Radeon     6450	Radeon     6570	Radeon     6670	Radeon     6750	Radeon     6770	Radeon     6790	Radeon     6850	Radeon     6870
Core
Codename	Caicos	Turks	Turks	Juniper	Juniper	Barts	Barts	Barts
Process scale (nm)	40	40	40	40	40	40	40	40
Transistor count (millions)	370	716	716	1040	1040	1700	1700	1700
Core clock speed (MHz)	750	650	800	700	850	840	775	900
Component execution units 1	160	480	480	720	800	800	960	1120
Vertex shader version (best)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Total effective no. of texture units	8	24	24	36	40	40	48	56
Max pixels output per clock (colour/depth)	4 / 16	8 / 32	8 / 32	16 / 64	16/ 64	16 / 64	32 / 128	32 / 128
Pixel shader version (bset)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Overall DirectX level	11	11	11	11	11	11	11	11
Memory
Interface width (bits)	64	128	128	128	128	256	256	256
Interface type	DDR3 / GDDR5	DDR3 / GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5
Clock speed (MHz)	800 / 900	900 / 1000	1000	1150	1200	1050	1000	1050
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	7.5	19.5	24.0	31.5	42.5	42.0	46.5	54.0
Triangle output rate (Gtris/sec)	0.75	0.65	0.80	0.72	0.85	0.80	0.78	0.90
Pixel processing rate (Gpixels/sec)	30.0	78.0	96.0	126.0	170.0	168.0	186.0	216.0
Texel applied rate (Gtexels/sec)	6.0	15.6	19.2	25.2	34.0	33.6	37.2	43.2
Pixel output rate (Gpixels/sec)	3.0	5.2	6.4	11.2	13.6	13.4	24.8	28.8
Memory bandwidth (GB/sec)	12.8 / 28.8	28.8 / 64.0	64.0	73.6	76.8	134.4	128.0	134.4

Model	ATI
	Radeon    6950	Radeon   6970	Radeon    6990
Core
Codename	Cayman	Cayman	Cayman
Process scale (nm)	40	40	40
Transistor count (millions)	2640	2640	2 x 2640
Core clock speed (MHz)	800	880	830
Component execution units 1	1408	1536	2 x 1536
Vertex shader version (best)	5.0	5.0	5.0
Total effective no. of texture units	88	96	2 x 96
Max pixels output per clock (colour/depth)	32 / 128	32 / 128	2 x 32 / 128
Pixel shader version (bset)	5.0	5.0	5.0
Overall DirectX level	11	11	11
Memory
Interface width (bits)	256	256	2 x 256
Interface type	GDDR5	GDDR5	GDDR5
Clock speed (MHz)	1250	1375	1250
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	70.4	84.5	159.4
Triangle output rate (Gtris/sec)	1.60	1.76	3.32
Pixel processing rate (Gpixels/sec)	281.6	337.9	637.4
Texel applied rate (Gtexels/sec)	70.4	84.5	159.4
Pixel output rate (Gpixels/sec)	25.6	28.2	53.1
Memory bandwidth (GB/sec)	160.0	176.0	320.0

Model	NVIDIA
	GeForce   GT 430	GeForce GTS 440	GeForce   GTS 450	GeForce   GTX 460 768 MiB	GeForce   GTX 460 1024 MiB	GeForce   GTX 465	GeForce GTX 470	GeForce GTX 480
Core
Codename	GF108	GF108	GF106	GF104	GF104	GF100	GF100	GF100
Process scale (nm)	40	40	40	40	40	40	40	40
Transistor count (millions)	585	585	~ 1170	~ 1950	~ 1950	~ 3000	~ 3000	~ 3000
Core / shader clock speed (MHz)	700 / 1400	810 / 1620	783 / 1577	675 / 1350	675 / 1350	607 / 1215	607 / 1215	700 / 1401
Component execution units 1	96	96	192	336	336	352	448	480
Vertex shader version (best)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Total effective no. of texture units	16 4	16 4	32 4	56 4	56 4	44 4	56 4	60 4
Max pixels output per clock (colour/depth)	4 / 32	4 / 32	16 / 128	24 / 256	32 / 192	32 / 256	40 / 320	48 / 384
Pixel shader version (bset)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Overall DirectX level	11	11	11	11	11	11	11	11
Memory
Interface width (bits)	128	128	128	192	256	256	320	384
Interface type	DDR3	GDDR3 / GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5
Clock speed (MHz)	900	900	900	900	900	802	837	924
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	8.4 3	9.7 3	37.8 3	56.8 3	56.8 3	53.5 3	68.1 3	84.1 3
Triangle output rate (Gtris/sec)	0.70	0.81	0.78	1.35	1.35	2.42	2.42	2.80
Pixel processing rate (Gpixels/sec)	33.6 3	38.9 3	151.4 3	226.8 3	226.8 3	213.9 3	272.2 3	336.2 3
Texel applied rate (Gtexels/sec)	11.2 4	13.0 4	25.2 4	37.8 4	37.8 4	26.7 4	34.0 4	42.0 4
Pixel output rate (Gpixels/sec)	2.8	3.2	12.5	16.2	21.6	19.4	24.3	33.6
Memory bandwidth (GB/sec)	28.8	28.8 / 57.6	57.6	86.4	115.2	102.7	133.9	177.4

Model	NVIDIA
	GeForce GT 520	GeForce GTX 550 Ti	GeForce GTX 560	GeForce   GTX 560 Ti	GeForce   GTX 560 Ti w448 cores	GeForce   GTX 570	GeForce   GTX 580	GeForce   GTX 590
Core
Codename	GF119	GF116	GF114	GF114	GF110	GF110	GF110	GF110
Process scale (nm)	40	40	40	40	40	40	40	40
Transistor count (millions)	?	~ 1170	~ 1950	~ 1950	~ 3000	~ 3000	~ 3000	2 x ~ 3000
Core / shader clock speed (MHz)	810 / 1620	900 / 1800	810 / 1620	822 / 1644	732 / 1464	732 / 1464	772 / 1544	607 / 1215
Component execution units 1	48	192	336	384	448	480	512	2 x 512
Vertex shader version (best)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Total effective no. of texture units	8 4	32 4	56 4	64 4	56 4	64 4	64 4	2 x 64 4
Max pixels output per clock (colour/depth)	4 / 32	24 / 192	32 / 192	32 / 192	32 / 192	40 / 320	48 / 384	2 x 48 / 384
Pixel shader version (bset)	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Overall DirectX level	11	11	11	11	11	11	11	11
Memory
Interface width (bits)	64	192	256	256	320	320	384	2 x 384
Interface type	GDDR3	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5
Clock speed (MHz)	900	1026	1002	1001	900	950	1002	854
Theoretical Peak Performance
Vertex processing rate (Gverts/sec)	9.72 3	43.2 3	68.0 3	78.9 3	81.9 3	87.8 3	98.2 3	155.5 2, 3
Triangle output rate (Gtris/sec)	0.81	1.64	1.64	1.64	2.73	2.93	3.09	4.86 2
Pixel processing rate (Gpixels/sec)	38.9 3	172.8 3	272.2 3	315.6 3	328.0 3	351.4 3	392.8 3	622.1 2, 3
Texel applied rate (Gtexels/sec)	6.5 4	25.9 4	45.4 4	52.6 4	41.0 4	46.8 4	49.4 4	77.7 2, 4
Pixel output rate (Gpixels/sec)	3.2	21.6	25.9	26.3	29.3	29.3	37.1	58.3 2
Memory bandwidth (GB/sec)	14.4	98.5	128.3	128.1	144.0	152.0	192.4	327.7 2

Notes

1 - All GPU vendors use a unified architecture for their DX11 chips - this means execution units are used to process all kinds of components (vertex, geometry and pixel).  Previous architectures had    pipelines of units dedicated for each type of component.

2 - This model is composed of two GPUs.  The performance figures are calculated for when operating in AFR (alternate frame rendering) mode.

3 - The GF1xx execution units contain a set of ALUs on two pipelines per unit.

4 - The GF1xx texture units run at half shader clock speed and can filter 4 times the number of texels that can be addressed.

3D Graphics Card Specifications - Copyright © - N D Evanson