GPU Name | Mars | GF108 |
---|---|---|
GPU Variante | Mars XT | GF108-300-A1 |
Architektur | GCN 1.0 | Fermi |
Hersteller | TSMC | TSMC |
Prozessgröße | 28 nm | 40 nm |
Transistoren | 950 million | 585 million |
Chipgröße | 77 mm² | 116 mm² |
Veröffentlichungsdatum | Mar 1st, 2013 | Sep 17th, 2012 |
---|---|---|
Generation | Solar System (HD 8600M) | NVS Mobile (x200M) |
Produktion | End-of-life | End-of-life |
Businterface | PCIe 3.0 x8 | MXM |
Vorgänger | London | — |
Nachfolger | Crystal System | — |
Reviews | — | 26 in our database |
Basis-Taktfrequenz | 775 MHz | — |
---|---|---|
Boost-Takt | 825 MHz | — |
Speicher-Taktung | 900 MHz 1800 Mbps effective | 785 MHz 3.1 Gbps effective |
GPU-Taktung | — | 672 MHz |
Shader-Taktung | — | 1344 MHz |
Speichergröße | 1024 MB | 1024 MB |
---|---|---|
Speichertyp | DDR3 | GDDR5 |
Speicherbus | 64 bit | 64 bit |
Bandbreite | 14.40 GB/s | 25.12 GB/s |
Shading-Einheiten | 384 | 96 |
---|---|---|
TMUs | 24 | 16 |
ROPs | 8 | 4 |
Compute Units | 6 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 128 KB | 128 KB |
Anzahl SM | — | 2 |
Pixelrate | 6.600 GPixel/s | 2.688 GPixel/s |
---|---|---|
Texturrate | 19.80 GTexel/s | 10.75 GTexel/s |
FP32 (float) Leistung | 633.6 GFLOPS | 258.0 GFLOPS |
FP64 (double) Leistung | — | 21.50 GFLOPS (1:12) |
TDP | unknown | 25 W |
---|---|---|
Ausgabe | No outputs | No outputs |
Steckplatzbreite | — | MXM Module |
DirectX | 12 (11_1) | 12 (11_0) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 1.2 | 1.1 |
Vulkan | 1.2 | — |
Shader-Modell | 5.1 | 5.1 |
CUDA | — | 2.1 |