Nom du GPU | Vega 10 | GM108 |
---|---|---|
GPU variant | Vega 10 XL (215-0894216) | — |
Architecture | GCN 5.0 | Maxwell |
Fonderie | GlobalFoundries | TSMC |
Taille du processus | 14 nm | 28 nm |
Transistors | 12,500 million | unknown |
Taille des matrices | 495 mm² | unknown |
Date de publication | Jun 1st, 2018 | Mar 12th, 2014 |
---|---|---|
Génération | Mobility Radeon (Vega) | GeForce 800M |
Production | Active | End-of-life |
Bus interface | PCIe 3.0 x16 | PCIe 3.0 x8 |
Critiques | 23 in our database | — |
Prédécesseur | Crystal System | GeForce 700M |
Successeur | — | GeForce 900M |
Base clock | 1138 MHz | 1082 MHz |
---|---|---|
Boost clock | 1301 MHz | 1150 MHz |
Mémoire clock | 800 MHz 1600 Mbps effective | 900 MHz 1800 Mbps effective |
Taille de la mémoire | 8 GB | 2 GB |
---|---|---|
Type de mémoire | HBM2 | DDR3 |
Bus mémoire | 2048 bit | 64 bit |
Bande passante | 409.6 GB/s | 14.40 GB/s |
Shading units | 3584 | 256 |
---|---|---|
TMUs | 224 | 16 |
ROPs | 64 | 8 |
Compute units | 56 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SMM) |
Cache L2 | 4 MB | 1024 KB |
Comptage SMM | — | 2 |
Taux de pixel | 83.26 GPixel/s | 9.200 GPixel/s |
---|---|---|
Taux de texture | 291.4 GTexel/s | 18.40 GTexel/s |
FP16 (half) performance | 18.65 TFLOPS (2:1) | — |
FP32 (float) performance | 9.326 TFLOPS | 588.8 GFLOPS |
FP64 (double) performance | 582.8 GFLOPS (1:16) | 18.40 GFLOPS (1:32) |
Taille Slot | MXM Module | — |
---|---|---|
Longueur | 105 mm 4.1 inches | — |
TDP | 120 W | 33 W |
Sorties | 1x HDMI3x DisplayPort | No outputs |
Connecteurs de puissance | None | — |
DirectX | 12 (12_1) | 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 |