GPU Name | Thames | Mars |
---|---|---|
GPU Variante | Thames XTX | Mars PRO (216-0842000) |
Architektur | TeraScale 2 | GCN 1.0 |
Hersteller | TSMC | TSMC |
Prozessgröße | 40 nm | 28 nm |
Transistoren | 716 million | 950 million |
Chipgröße | 104 mm² | 77 mm² |
Veröffentlichungsdatum | Jan 7th, 2013 | Feb 26th, 2013 |
---|---|---|
Generation | London (HD 7600M) | Solar System (HD 8700M) |
Produktion | End-of-life | End-of-life |
Businterface | PCIe 2.0 x16 | PCIe 3.0 x8 |
Vorgänger | Vancouver | London |
Nachfolger | Solar System | Crystal System |
GPU-Taktung | 725 MHz | — |
---|---|---|
Speicher-Taktung | 900 MHz 3.6 Gbps effective | 1000 MHz 4 Gbps effective |
Basis-Taktfrequenz | — | 620 MHz |
Boost-Takt | — | 775 MHz |
Speichergröße | 2 GB | 1024 MB |
---|---|---|
Speichertyp | GDDR5 | GDDR5 |
Speicherbus | 128 bit | 128 bit |
Bandbreite | 57.60 GB/s | 64.00 GB/s |
Shading-Einheiten | 480 | 384 |
---|---|---|
TMUs | 24 | 24 |
ROPs | 8 | 8 |
Compute Units | 6 | 6 |
Cache L1 | 8 KB (per CU) | 16 KB (per CU) |
Cache L2 | 256 KB | 256 KB |
Pixelrate | 5.800 GPixel/s | 6.200 GPixel/s |
---|---|---|
Texturrate | 17.40 GTexel/s | 18.60 GTexel/s |
FP32 (float) Leistung | 696.0 GFLOPS | 595.2 GFLOPS |
FP64 (double) Leistung | — | 37.20 GFLOPS (1:16) |
TDP | 25 W | unknown |
---|---|---|
Ausgabe | No outputs | No outputs |
DirectX | 11.2 (11_0) | 12 (11_1) |
---|---|---|
OpenGL | 4.4 | 4.6 |
OpenCL | 1.2 | 1.2 |
Vulkan | — | 1.2 |
Shader-Modell | 5.0 | 5.1 |