Nom du GPU | GM108 | GV10B |
---|---|---|
Architecture | Maxwell | Volta |
Fonderie | TSMC | TSMC |
Taille du processus | 28 nm | 12 nm |
Transistors | unknown | 9,000 million |
Taille des matrices | unknown | 350 mm² |
Date de publication | Jul 22nd, 2014 | — |
---|---|---|
Génération | GeForce 800A | — |
Production | End-of-life | — |
Bus interface | PCIe 3.0 x8 | — |
Prédécesseur | GeForce 700A | — |
Successeur | GeForce 900A | — |
Base clock | 1082 MHz | 854 MHz |
---|---|---|
Boost clock | 1150 MHz | 1377 MHz |
Mémoire clock | 900 MHz 1800 Mbps effective | System Shared |
Taille de la mémoire | 2 GB | System Shared |
---|---|---|
Type de mémoire | DDR3 | System Shared |
Bus mémoire | 64 bit | System Shared |
Bande passante | 14.40 GB/s | System Dependent |
Shading units | 256 | 512 |
---|---|---|
TMUs | 16 | 32 |
ROPs | 8 | 16 |
Comptage SMM | 2 | — |
Cache L1 | 64 KB (per SMM) | 128 KB (per SM) |
Cache L2 | 1024 KB | 512 KB |
Comptage SM | — | 8 |
Tensor cores | — | 64 |
Taux de pixel | 9.200 GPixel/s | 22.03 GPixel/s |
---|---|---|
Taux de texture | 18.40 GTexel/s | 44.06 GTexel/s |
FP32 (float) performance | 588.8 GFLOPS | 1,410 GFLOPS |
FP64 (double) performance | 18.40 GFLOPS (1:32) | 705.0 GFLOPS (1:2) |
FP16 (half) performance | — | 2.820 TFLOPS (2:1) |
Taille Slot | IGP | IGP |
---|---|---|
TDP | 33 W | 30 W |
Sorties | No outputs | No outputs |
Longueur | — | 100 mm 3.9 inches |
Taille | — | 87 mm 3.4 inches |
DirectX | 12 (11_0) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 1.2 |
Vulkan | 1.1 | 1.2 |
CUDA | 5.0 | 7.2 |
Shader model | 5.1 | 6.4 |
Date de publication | — | Oct 2018 |
---|---|---|
Génération | — | Tegra |
Production | — | End-of-life |
Prix de lancement | — | 899 USD |
Bus interface | — | IGP |