GPU model | M10 | GF108 |
---|---|---|
Architecture | Rage 8 | Fermi |
Foundry | TSMC | TSMC |
Manufacturing process | 130 nm | 40 nm |
Number of transistors | 76 million | 585 million |
Die size | 92 mm² | 116 mm² |
GPU variant | — | GF108-200-A1 |
Release date | Jun 1st, 2004 | — |
---|---|---|
Generation | M1x (Mobility 9000) | — |
Production | End-of-life | — |
Interface | AGP 8x | — |
Reviews | 2 in our database | — |
Predecessor | M9 | — |
Successor | M2x | — |
GPU clock | 333 MHz | 700 MHz |
---|---|---|
Memory clock | 200 MHz 400 Mbps effective | 900 MHz 1800 Mbps effective |
Shader clock | — | 1400 MHz |
Memory size | 128 MB | 1024 MB |
---|---|---|
Memory type | DDR | DDR3 |
Memory bus width | 128 bit | 128 bit |
Bandwidth | 6.400 GB/s | 28.80 GB/s |
Pixel shaders | 4 | — |
---|---|---|
Vertex shaders | 2 | — |
TMUs | 4 | 4 |
ROPs | 4 | 4 |
Shading units | — | 48 |
SM count | — | 1 |
Cache L1 | — | 64 KB (per SM) |
Cache L2 | — | 256 KB |
Pixel rate | 1.332 GPixel/s | 1.400 GPixel/s |
---|---|---|
Vertex rate | 166.5 MVertices/s | — |
Texture rate | 1.332 GTexel/s | 2.800 GTexel/s |
FP32 (float) performance | — | 134.4 GFLOPS |
FP64 (double) performance | — | 11.20 GFLOPS (1:12) |
TDP | unknown | 50 W |
---|---|---|
Outputs | No outputs | 1x DVI1x HDMI1x VGA |
Slot width | — | Single-slot |
Length | — | 145 mm 5.7 inches |
Suggested PSU | — | 250 W |
Power connectors | — | None |
Board number | — | P1071 |
DirectX | 9.0 (9_0) | 12 (11_0) |
---|---|---|
OpenGL | 2.0 | 4.6 |
OpenCL | — | 1.1 |
Vulkan | — | — |
Pixel shader | 2.0 | — |
Vertex shader | 2.0 | — |
CUDA | — | 2.1 |
Shader model | — | 5.1 |
Release date | — | Sep 3rd, 2010 |
---|---|---|
Generation | — | GeForce 400 |
Production | — | End-of-life |
Interface | — | PCIe 2.0 x16 |
Predecessor | — | GeForce 200 |
Successor | — | GeForce 500 |