Nom du GPU | Polaris 21 | GP100 |
---|---|---|
GPU variant | Polaris 21 XT (215-0908004) | GP100-893-A1 |
Architecture | GCN 4.0 | Pascal |
Fonderie | GlobalFoundries | TSMC |
Taille du processus | 14 nm | 16 nm |
Transistors | 3,000 million | 15,300 million |
Taille des matrices | 123 mm² | 610 mm² |
Date de publication | Apr 11th, 2018 | Jun 20th, 2016 |
---|---|---|
Génération | Polaris | Tesla |
Production | Active | End-of-life |
Bus interface | PCIe 3.0 x8 | PCIe 3.0 x16 |
Prédécesseur | Polaris | — |
Successeur | Vega | — |
Prix de lancement | — | 5,699 USD |
Base clock | 1175 MHz | 1190 MHz |
---|---|---|
Boost clock | 1275 MHz | 1329 MHz |
Mémoire clock | 1750 MHz 7 Gbps effective | 715 MHz 1430 Mbps effective |
Taille de la mémoire | 4 GB | 16 GB |
---|---|---|
Type de mémoire | GDDR5 | HBM2 |
Bus mémoire | 128 bit | 4096 bit |
Bande passante | 112.0 GB/s | 732.2 GB/s |
Shading units | 1024 | 3584 |
---|---|---|
TMUs | 64 | 224 |
ROPs | 16 | 96 |
Compute units | 16 | — |
Cache L1 | 16 KB (per CU) | 24 KB (per SM) |
Cache L2 | 1024 KB | 4 MB |
Comptage SM | — | 56 |
Taux de pixel | 20.40 GPixel/s | 127.6 GPixel/s |
---|---|---|
Taux de texture | 81.60 GTexel/s | 297.7 GTexel/s |
FP16 (half) performance | 2.611 TFLOPS (1:1) | 19.05 TFLOPS (2:1) |
FP32 (float) performance | 2.611 TFLOPS | 9.526 TFLOPS |
FP64 (double) performance | 163.2 GFLOPS (1:16) | 4.763 TFLOPS (1:2) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 170 mm 6.7 inches | 267 mm 10.5 inches |
TDP | 75 W | 250 W |
Suggestions pour le PSU | 250 W | 600 W |
Sorties | 1x DVI1x HDMI1x DisplayPort | No outputs |
Connecteurs de puissance | None | 1x 8-pin |
Board number | C994 | — |
DirectX | 12 (12_0) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.4 | 6.4 |
CUDA | — | 6.0 |