Nom du GPU | TU106 | GP100 |
---|---|---|
GPU variant | TU106-410-A1 | GP100-893-A1 |
Architecture | Turing | Pascal |
Fonderie | TSMC | TSMC |
Taille du processus | 12 nm | 16 nm |
Transistors | 10,800 million | 15,300 million |
Taille des matrices | 445 mm² | 610 mm² |
Date de publication | Jul 9th, 2019 | Jun 20th, 2016 |
---|---|---|
Génération | GeForce 20 | Tesla |
Prédécesseur | GeForce 10 | — |
Production | Active | End-of-life |
Prix de lancement | 399 USD | 5,699 USD |
Bus interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Critiques | 40 in our database | — |
Successeur | GeForce 30 | — |
Base clock | 1470 MHz | 1190 MHz |
---|---|---|
Boost clock | 1650 MHz | 1329 MHz |
Mémoire clock | 1750 MHz 14 Gbps effective | 715 MHz 1430 Mbps effective |
Taille de la mémoire | 8 GB | 16 GB |
---|---|---|
Type de mémoire | GDDR6 | HBM2 |
Bus mémoire | 256 bit | 4096 bit |
Bande passante | 448.0 GB/s | 732.2 GB/s |
Shading units | 2176 | 3584 |
---|---|---|
TMUs | 136 | 224 |
ROPs | 64 | 96 |
Comptage SM | 34 | 56 |
Tensor cores | 272 | — |
RT cores | 34 | — |
Cache L1 | 64 KB (per SM) | 24 KB (per SM) |
Cache L2 | 4 MB | 4 MB |
Taux de pixel | 105.6 GPixel/s | 127.6 GPixel/s |
---|---|---|
Taux de texture | 224.4 GTexel/s | 297.7 GTexel/s |
FP16 (half) performance | 14.36 TFLOPS (2:1) | 19.05 TFLOPS (2:1) |
FP32 (float) performance | 7.181 TFLOPS | 9.526 TFLOPS |
FP64 (double) performance | 224.4 GFLOPS (1:32) | 4.763 TFLOPS (1:2) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 229 mm 9 inches | 267 mm 10.5 inches |
Taille | 113 mm 4.4 inches | — |
Taille | 35 mm 1.4 inches | — |
TDP | 175 W | 250 W |
Suggestions pour le PSU | 450 W | 600 W |
Sorties | 1x DVI1x HDMI2x DisplayPort1x USB Type-C | No outputs |
Connecteurs de puissance | 1x 8-pin | 1x 8-pin |
Board number | PG160 SKU 65 | — |
DirectX | 12 Ultimate (12_2) | 12 (12_1) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 7.5 | 6.0 |
Shader model | 6.6 | 6.4 |