GPU Name | GP104 | GA106 |
---|---|---|
Architektur | Pascal | Ampere |
Hersteller | TSMC | Samsung |
Prozessgröße | 16 nm | 8 nm |
Transistoren | 7,200 million | 13,250 million |
Chipgröße | 314 mm² | 276 mm² |
Veröffentlichungsdatum | Feb 21st, 2018 | Apr 12th, 2021 |
---|---|---|
Generation | Quadro Mobile (Px200) | Quadro Mobile (Ax000) |
Produktion | End-of-life | Active |
Businterface | MXM-B (3.0) | PCIe 4.0 x16 |
Basis-Taktfrequenz | 1316 MHz | 893 MHz |
---|---|---|
Boost-Takt | 1569 MHz | 1358 MHz |
Speicher-Taktung | 1804 MHz 7.2 Gbps effective | 1375 MHz 11 Gbps effective |
Speichergröße | 16 GB | 4 GB |
---|---|---|
Speichertyp | GDDR5 | GDDR6 |
Speicherbus | 256 bit | 128 bit |
Bandbreite | 230.9 GB/s | 176.0 GB/s |
Shading-Einheiten | 2560 | 2560 |
---|---|---|
TMUs | 160 | 80 |
ROPs | 64 | 48 |
Anzahl SM | 20 | 20 |
Cache L1 | 48 KB (per SM) | 128 KB (per SM) |
Cache L2 | 2 MB | 2 MB |
Tensor Cores | — | 80 |
RT-Kerne | — | 20 |
Pixelrate | 100.4 GPixel/s | 65.18 GPixel/s |
---|---|---|
Texturrate | 251.0 GTexel/s | 108.6 GTexel/s |
FP16 (half) Leistung | 125.5 GFLOPS (1:64) | 6.953 TFLOPS (1:1) |
FP32 (float) Leistung | 8.033 TFLOPS | 6.953 TFLOPS |
FP64 (double) Leistung | 251.0 GFLOPS (1:32) | 108.6 GFLOPS (1:64) |
Steckplatzbreite | MXM Module | IGP |
---|---|---|
TDP | 100 W | 95 W |
Ausgabe | No outputs | No outputs |
Stromverbindung | None | None |
DirectX | 12 (12_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 6.1 | 8.6 |
Shader-Modell | 6.4 | 6.6 |