GPU model | Vega 20 | TU106 |
---|---|---|
GPU variant | Vega 20 GLXT | N18E-G2-A1 |
Architecture | GCN 5.1 | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 7 nm | 12 nm |
Number of transistors | 13,230 million | 10,800 million |
Die size | 331 mm² | 445 mm² |
Release date | Nov 18th, 2018 | — |
---|---|---|
Generation | Radeon Instinct | — |
Production | Active | — |
Interface | PCIe 4.0 x16 | — |
Base clock | 1200 MHz | 885 MHz |
---|---|---|
Boost clock | 1746 MHz | 1185 MHz |
Memory clock | 1000 MHz 2 Gbps effective | 1500 MHz 12 Gbps effective |
Memory size | 16 GB | 8 GB |
---|---|---|
Memory type | HBM2 | GDDR6 |
Memory bus width | 4096 bit | 256 bit |
Bandwidth | 1,024 GB/s | 384.0 GB/s |
Shading units | 3840 | 2304 |
---|---|---|
TMUs | 240 | 144 |
ROPs | 64 | 64 |
Compute units | 60 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 4 MB | 4 MB |
SM count | — | 36 |
Tensor cores | — | 288 |
RT cores | — | 36 |
Pixel rate | 111.7 GPixel/s | 75.84 GPixel/s |
---|---|---|
Texture rate | 419.0 GTexel/s | 170.6 GTexel/s |
FP16 (half) performance | 26.82 TFLOPS (2:1) | 10.92 TFLOPS (2:1) |
FP32 (float) performance | 13.41 TFLOPS | 5.460 TFLOPS |
FP64 (double) performance | 6.705 TFLOPS (1:2) | 170.6 GFLOPS (1:32) |
Slot width | Dual-slot | MXM Module |
---|---|---|
Length | 267 mm 10.5 inches | — |
TDP | 300 W | 90 W |
Suggested PSU | 700 W | — |
Outputs | 1x mini-DisplayPort | No outputs |
Power connectors | 1x 6-pin + 1x 8-pin | None |
Board number | — | E4914 SKU 30 |
DirectX | 12 (12_1) | 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 |
Release date | — | Jan 29th, 2019 |
---|---|---|
Generation | — | GeForce 20 Mobile |
Production | — | Active |
Interface | — | PCIe 3.0 x16 |
Predecessor | — | GeForce 10 Mobile |
Successor | — | GeForce 30 Mobile |