Nom du GPU | Ellesmere | TU104 |
---|---|---|
GPU variant | Ellesmere XT | — |
Architecture | GCN 4.0 | Turing |
Fonderie | GlobalFoundries | TSMC |
Taille du processus | 14 nm | 12 nm |
Transistors | 5,700 million | 13,600 million |
Taille des matrices | 232 mm² | 545 mm² |
Date de publication | Aug 4th, 2016 | May 27th, 2019 |
---|---|---|
Génération | Mobility Radeon (RX M400) | Quadro Mobile (Tx000) |
Production | Active | Active |
Bus interface | MXM-B (3.0) | PCIe 3.0 x16 |
Prédécesseur | Crystal System | — |
Base clock | 1000 MHz | 600 MHz |
---|---|---|
Boost clock | 1077 MHz | 1350 MHz |
Mémoire clock | 2000 MHz 8 Gbps effective | 1500 MHz 12 Gbps effective |
Taille de la mémoire | 8 GB | 16 GB |
---|---|---|
Type de mémoire | GDDR5 | GDDR6 |
Bus mémoire | 256 bit | 256 bit |
Bande passante | 256.0 GB/s | 384.0 GB/s |
Shading units | 2304 | 3072 |
---|---|---|
TMUs | 144 | 192 |
ROPs | 32 | 64 |
Compute units | 36 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 2 MB | 4 MB |
Comptage SM | — | 48 |
Tensor cores | — | 384 |
RT cores | — | 48 |
Taux de pixel | 34.46 GPixel/s | 86.40 GPixel/s |
---|---|---|
Taux de texture | 155.1 GTexel/s | 259.2 GTexel/s |
FP16 (half) performance | 4.963 TFLOPS (1:1) | 16.59 TFLOPS (2:1) |
FP32 (float) performance | 4.963 TFLOPS | 8.294 TFLOPS |
FP64 (double) performance | 310.2 GFLOPS (1:16) | 259.2 GFLOPS (1:32) |
Taille Slot | MXM Module | IGP |
---|---|---|
TDP | 100 W | 80 W |
Sorties | No outputs | No outputs |
Connecteurs de puissance | None | None |
DirectX | 12 (12_0) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.4 | 6.6 |
CUDA | — | 7.5 |