Nom du GPU | Baffin | GM107 |
---|---|---|
GPU variant | Baffin LE | GM107-570-A2 |
Architecture | GCN 4.0 | Maxwell |
Fonderie | GlobalFoundries | TSMC |
Taille du processus | 14 nm | 28 nm |
Transistors | 3,000 million | 1,870 million |
Taille des matrices | 123 mm² | 148 mm² |
Date de publication | Oct 13th, 2017 | May 18th, 2016 |
---|---|---|
Génération | Polaris | GRID |
Production | Active | End-of-life |
Bus interface | PCIe 3.0 x8 | PCIe 3.0 x16 |
Prédécesseur | Arctic Islands | — |
Successeur | Vega | — |
Base clock | 1019 MHz | 1033 MHz |
---|---|---|
Boost clock | 1071 MHz | 1306 MHz |
Mémoire clock | 1500 MHz 6 Gbps effective | 1300 MHz 5.2 Gbps effective |
Taille de la mémoire | 2 GB | 8 GB |
---|---|---|
Type de mémoire | GDDR5 | GDDR5 |
Bus mémoire | 128 bit | 128 bit |
Bande passante | 96.00 GB/s | 83.20 GB/s |
Shading units | 640 | 640 |
---|---|---|
TMUs | 40 | 40 |
ROPs | 16 | 16 |
Compute units | 10 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SMM) |
Cache L2 | 1024 KB | 2 MB |
Comptage SMM | — | 5 |
Taux de pixel | 17.14 GPixel/s | 20.90 GPixel/s |
---|---|---|
Taux de texture | 42.84 GTexel/s | 52.24 GTexel/s |
FP16 (half) performance | 1,371 GFLOPS (1:1) | — |
FP32 (float) performance | 1,371 GFLOPS | 1.672 TFLOPS |
FP64 (double) performance | 85.68 GFLOPS (1:16) | 52.24 GFLOPS (1:32) |
Taille Slot | Dual-slot | Dual-slot |
---|---|---|
Longueur | 145 mm 5.7 inches | 267 mm 10.5 inches |
TDP | 60 W | 225 W |
Suggestions pour le PSU | 250 W | 550 W |
Sorties | 1x DVI1x HDMI1x DisplayPort | No outputs |
Connecteurs de puissance | None | 1x 8-pin |
Board number | D090-21 | — |
DirectX | 12 (12_0) | 12 (11_0) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.1 |
Shader model | 6.4 | 5.1 |
CUDA | — | 5.0 |