GPU model | M97 | GF108 |
---|---|---|
Architecture | TeraScale | Fermi |
Foundry | TSMC | TSMC |
Manufacturing process | 40 nm | 40 nm |
Number of transistors | 826 million | 585 million |
Die size | 137 mm² | 116 mm² |
GPU variant | — | GF108-400-A1 |
Release date | Mar 3rd, 2009 | — |
---|---|---|
Generation | M9x (Mobility HD 4800) | — |
Production | End-of-life | — |
Interface | PCIe 2.0 x16 | — |
Reviews | 2 in our database | — |
Predecessor | M8x | — |
Successor | Manhattan | — |
GPU clock | 550 MHz | 700 MHz |
---|---|---|
Memory clock | 700 MHz 2.8 Gbps effective | 800 MHz 1600 Mbps effective |
Shader clock | — | 1400 MHz |
Memory size | 1024 MB | 2 GB |
---|---|---|
Memory type | GDDR5 | DDR3 |
Memory bus width | 128 bit | 128 bit |
Bandwidth | 44.80 GB/s | 25.60 GB/s |
Shading units | 640 | 96 |
---|---|---|
TMUs | 32 | 16 |
ROPs | 16 | 4 |
Compute units | 8 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 128 KB | 256 KB |
SM count | — | 2 |
Pixel rate | 8.800 GPixel/s | 2.800 GPixel/s |
---|---|---|
Texture rate | 17.60 GTexel/s | 11.20 GTexel/s |
FP32 (float) performance | 704.0 GFLOPS | 268.8 GFLOPS |
FP64 (double) performance | 140.8 GFLOPS (1:5) | 22.40 GFLOPS (1:12) |
TDP | unknown | 49 W |
---|---|---|
Outputs | No outputs | 1x DVI1x HDMI1x VGA |
Slot width | — | Single-slot |
Length | — | 145 mm 5.7 inches |
Suggested PSU | — | 200 W |
Power connectors | — | None |
Board number | — | P1070, P1071 |
DirectX | 10.1 (10_1) | 12 (11_0) |
---|---|---|
OpenGL | 3.3 | 4.6 |
OpenCL | 1.1 | 1.1 |
Vulkan | — | — |
Shader model | 4.1 | 5.1 |
CUDA | — | 2.1 |
Release date | — | Jun 18th, 2014 |
---|---|---|
Generation | — | GeForce 700 |
Production | — | End-of-life |
Interface | — | PCIe 2.0 x16 |
Predecessor | — | GeForce 600 |
Successor | — | GeForce 900 |