Nom du GPU | GP106 | TU104 |
---|---|---|
GPU variant | GP106-875-A1 | TU104-850-A1 |
Architecture | Pascal | Turing |
Fonderie | TSMC | TSMC |
Taille du processus | 16 nm | 12 nm |
Transistors | 4,400 million | 13,600 million |
Taille des matrices | 200 mm² | 545 mm² |
Date de publication | Feb 6th, 2017 | Nov 13th, 2018 |
---|---|---|
Génération | Quadro | Quadro |
Production | End-of-life | Active |
Bus interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Prix de lancement | — | 899 USD |
Base clock | 1076 MHz | 1005 MHz |
---|---|---|
Boost clock | 1480 MHz | 1545 MHz |
Mémoire clock | 1752 MHz 7 Gbps effective | 1625 MHz 13 Gbps effective |
Taille de la mémoire | 5 GB | 8 GB |
---|---|---|
Type de mémoire | GDDR5 | GDDR6 |
Bus mémoire | 160 bit | 256 bit |
Bande passante | 140.2 GB/s | 416.0 GB/s |
Shading units | 1024 | 2304 |
---|---|---|
TMUs | 64 | 144 |
ROPs | 40 | 64 |
Comptage SM | 8 | 36 |
Cache L1 | 48 KB (per SM) | 64 KB (per SM) |
Cache L2 | 1280 KB | 4 MB |
Tensor cores | — | 288 |
RT cores | — | 36 |
Taux de pixel | 59.20 GPixel/s | 98.88 GPixel/s |
---|---|---|
Taux de texture | 94.72 GTexel/s | 222.5 GTexel/s |
FP16 (half) performance | 47.36 GFLOPS (1:64) | 14.24 TFLOPS (2:1) |
FP32 (float) performance | 3.031 TFLOPS | 7.119 TFLOPS |
FP64 (double) performance | 94.72 GFLOPS (1:32) | 222.5 GFLOPS (1:32) |
Taille Slot | Single-slot | Single-slot |
---|---|---|
Longueur | 196 mm 7.7 inches | 241 mm 9.5 inches |
Taille | 111 mm 4.4 inches | 111 mm 4.4 inches |
TDP | 75 W | 160 W |
Suggestions pour le PSU | 250 W | 450 W |
Sorties | 4x DisplayPort | 3x DisplayPort1x USB Type-C |
Connecteurs de puissance | None | 1x 8-pin |
Board number | PG410 SKU 502 | — |
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 | 7.5 |
Shader model | 6.4 | 6.6 |