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    
5570
    Radeon    
5670
    Radeon    
5750
    Radeon    
5770
    Radeon    
5830
    Radeon    
5850
    Radeon    
5870
 Core
  CodenameCedarRedwoodRedwoodJuniperJuniperCypressCypressCypress
  Process scale (nm) 4040404040 40 40 40
  Transistor count (millions) 29262762710401040 2150 2150 2150
  Core clock speed (MHz) 650650775700850 800 725 850
  Component execution units 1 80400400720800 1120 1440 1600
  Vertex shader version (best) 5.05.05.05.05.0 5.0 5.0 5.0
  Total effective no. of texture units 820203240 56 72 80
  Max pixels output per clock (colour/depth) 4 / 168 / 328 / 3216 / 6416 / 64 16 / 64 32 / 128 32 / 128
  Pixel shader version (bset) 5.05.05.05.05.0 5.0 5.0 5.0
  Overall DirectX level 1111111111 11 11 11
 Memory
  Interface width (bits) 64128128128128 256 256 256
  Interface type DDR2 / DDR3DDR3GDDR5GDDR5GDDR5 GDDR5 GDDR5 GDDR5
  Clock speed (MHz) 800900100011501200 1000 1000 1200
 Theoretical Peak Performance
  Vertex processing rate (Gverts/sec) 3.2516.319.431.542.5 56.0 65.3 85.0
  Pixel processing rate (Gpixels/sec) 13.065.077.5126.0170.0 224.0 261.0 340.0
  Texel applied rate (Gtexels/sec) 5.213.015.528.034.0 44.8 52.2 68.0
  Pixel output rate (Gpixels/sec) 2.65.26.211.213.6 12.8 23.2 27.2
  Memory bandwidth (GB/sec) 6.428.864.073.676.8 128.0 128.0 153.6
  
 Model         ATI
    Radeon    
5890
                                                                                                            
 Core
  CodenameHemlock
  Process scale (nm) 40
  Transistor count (millions) 2 x 2150
  Core clock speed (MHz) 725
  Component execution units 1 2 x 1600
  Vertex shader version (best) 5.0
  Total effective no. of texture units 2 x 80
  Max pixels output per clock (colour/depth) 2 x 32 / 128
  Pixel shader version (bset) 5.0
  Overall DirectX level 11
 Memory
  Interface width (bits) 2 x 256
  Interface type GDDR5
  Clock speed (MHz) 1000
 Theoretical Peak Performance
  Vertex processing rate (Gverts/sec) 145.0 2
  Pixel processing rate (Gpixels/sec) 580.0 2
  Texel applied rate (Gtexels/sec) 116.0 2
  Pixel output rate (Gpixels/sec) 46.4 2
  Memory bandwidth (GB/sec) 256.0 2
  
 


 Model      NVIDIA
   GeForce  
GTX 460
768 MiB
   GeForce  
GTX 460
1024 MiB
   GeForce  
GTX 465
GeForce
GTX 470
   GeForce  
   GTX 475   
GeForce
GTX 480
                         
 Core
  CodenameGF104GF104GF100GF100GF104GF100
  Process scale (nm) 40404040 40 40
  Transistor count (millions) ~ 1950~ 1950~ 3000~ 3000 ~ 1950 ~ 3000
  Core / shader clock speed (MHz) 675 / 1350675 / 1350607 / 1215607 / 1215 ? / ? 700 / 1401
  Component execution units 1 336336352448 ? 480
  Vertex shader version (best) 5.05.05.05.0 5.0 5.0
  Total effective no. of texture units 56 456 444 456 4 4 60 4
  Max pixels output per clock (colour/depth) 24 / 25632 / 19232 / 25640 / 320 ? / ? 48 / 384
  Pixel shader version (bset) 5.05.05.05.0 5.0 5.0
  Overall DirectX level 11111111 11 11
 Memory
  Interface width (bits) 192256256320 ? 384
  Interface type GDDR5GDDR5GDDR5GDDR5 GDDR5 GDDR5
  Clock speed (MHz) 900900802837 ? 924
 Theoretical Peak Performance
  Vertex processing rate (Gverts/sec) 56.8 356.8 353.5 368.1 3 3 84.1 3
  Pixel processing rate (Gpixels/sec) 226.8 3226.8 3213.9 3272.2 3 3 336.2 3
  Texel applied rate (Gtexels/sec) 37.8 437.8 426.7 434.0 4 4 42.0 4
  Pixel output rate (Gpixels/sec) 16.221.619.424.3 ? 33.6
  Memory bandwidth (GB/sec) 86.4115.2102.7133.9 ? 177.4
  
 


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 - The ATI Radeon HD 5890 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 - N D Evanson 2010