Nom du GPU | GP100 | GP102 |
---|---|---|
Architecture | Pascal | Pascal |
Fonderie | TSMC | TSMC |
Taille du processus | 16 nm | 16 nm |
Transistors | 15,300 million | 11,800 million |
Taille des matrices | 610 mm² | 471 mm² |
Date de publication | Jun 20th, 2016 | Sep 13th, 2016 |
---|---|---|
Génération | Tesla | Tesla |
Production | End-of-life | End-of-life |
Prix de lancement | 4,599 USD | 5,699 USD |
Bus interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Base clock | 1190 MHz | 1303 MHz |
---|---|---|
Boost clock | 1329 MHz | 1531 MHz |
Mémoire clock | 715 MHz 1430 Mbps effective | 1808 MHz 14.5 Gbps effective |
Taille de la mémoire | 12 GB | 24 GB |
---|---|---|
Type de mémoire | HBM2 | GDDR5X |
Bus mémoire | 3072 bit | 384 bit |
Bande passante | 549.1 GB/s | 694.3 GB/s |
Shading units | 3584 | 3840 |
---|---|---|
TMUs | 224 | 240 |
ROPs | 96 | 96 |
Comptage SM | 56 | 30 |
Cache L1 | 24 KB (per SM) | 48 KB (per SM) |
Cache L2 | 3 MB | 3 MB |
Taux de pixel | 127.6 GPixel/s | 147.0 GPixel/s |
---|---|---|
Taux de texture | 297.7 GTexel/s | 367.4 GTexel/s |
FP16 (half) performance | 19.05 TFLOPS (2:1) | 183.7 GFLOPS (1:64) |
FP32 (float) performance | 9.526 TFLOPS | 11.76 TFLOPS |
FP64 (double) performance | 4.763 TFLOPS (1:2) | 367.4 GFLOPS (1:32) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 267 mm 10.5 inches | 267 mm 10.5 inches |
TDP | 250 W | 250 W |
Suggestions pour le PSU | 600 W | 600 W |
Sorties | No outputs | No outputs |
Connecteurs de puissance | 1x 8-pin | 1x 6-pin + 1x 8-pin |
Taille | — | 111 mm 4.4 inches |
Board number | — | PG610 SKU 200 |
DirectX | 12 (12_1) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 6.0 | 6.1 |
Shader model | 6.4 | 6.4 |