Nom du GPU | Curacao | TU102 |
---|---|---|
GPU variant | Curacao XT (215-0848004) | TU102-875-A1 |
Architecture | GCN 1.0 | Turing |
Fonderie | TSMC | TSMC |
Taille du processus | 28 nm | 12 nm |
Transistors | 2,800 million | 18,600 million |
Taille des matrices | 212 mm² | 754 mm² |
Date de publication | Oct 8th, 2013 | Aug 13th, 2018 |
---|---|---|
Génération | Volcanic Islands | Quadro |
Production | End-of-life | Active |
Prix de lancement | 199 USD | 6,299 USD |
Bus interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Critiques | 90 in our database | — |
Prédécesseur | Sea Islands | — |
Successeur | Pirate Islands | — |
Base clock | 1000 MHz | 1440 MHz |
---|---|---|
Boost clock | 1050 MHz | 1770 MHz |
Mémoire clock | 1400 MHz 5.6 Gbps effective | 1750 MHz 14 Gbps effective |
Taille de la mémoire | 2 GB | 24 GB |
---|---|---|
Type de mémoire | GDDR5 | GDDR6 |
Bus mémoire | 256 bit | 384 bit |
Bande passante | 179.2 GB/s | 672.0 GB/s |
Shading units | 1280 | 4608 |
---|---|---|
TMUs | 80 | 288 |
ROPs | 32 | 96 |
Compute units | 20 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 512 KB | 6 MB |
Comptage SM | — | 72 |
Tensor cores | — | 576 |
RT cores | — | 72 |
Taux de pixel | 33.60 GPixel/s | 169.9 GPixel/s |
---|---|---|
Taux de texture | 84.00 GTexel/s | 509.8 GTexel/s |
FP32 (float) performance | 2.688 TFLOPS | 16.31 TFLOPS |
FP64 (double) performance | 168.0 GFLOPS (1:16) | 509.8 GFLOPS (1:32) |
FP16 (half) performance | — | 32.62 TFLOPS (2:1) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 241 mm 9.5 inches | 267 mm 10.5 inches |
Taille | 109 mm 4.3 inches | 111 mm 4.4 inches |
Taille | 36 mm 1.4 inches | — |
TDP | 180 W | 260 W |
Suggestions pour le PSU | 450 W | 600 W |
Sorties | 2x DVI1x HDMI1x DisplayPort | 4x DisplayPort1x USB Type-C |
Connecteurs de puissance | 2x 6-pin | 1x 6-pin + 1x 8-pin |
Board number | C631, C632 | PG150 SKU 510 |
DirectX | 12 (11_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 1.2 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 5.1 | 6.6 |
CUDA | — | 7.5 |