GPU model | Raven | TU106 |
---|---|---|
Architecture | GCN 5.0 | Turing |
Foundry | GlobalFoundries | TSMC |
Manufacturing process | 14 nm | 12 nm |
Number of transistors | 4,940 million | 10,800 million |
Die size | 210 mm² | 445 mm² |
GPU variant | — | TU106-300-KA-A1 |
Release date | Apr 19th, 2018 | — |
---|---|---|
Generation | Raven Ridge (Vega) | — |
Production | Active | — |
Interface | IGP | — |
Predecessor | Great Horned Owl | — |
Successor | Picasso | — |
Base clock | 300 MHz | 1470 MHz |
---|---|---|
Boost clock | 1101 MHz | 1650 MHz |
Memory clock | System Shared | 1750 MHz 14 Gbps effective |
Memory size | System Shared | 12 GB |
---|---|---|
Memory type | System Shared | GDDR6 |
Memory bus width | System Shared | 192 bit |
Bandwidth | System Dependent | 336.0 GB/s |
Shading units | 512 | 2176 |
---|---|---|
TMUs | 32 | 136 |
ROPs | 8 | 48 |
Compute units | 8 | — |
SM count | — | 34 |
Tensor cores | — | 272 |
RT cores | — | 34 |
Cache L1 | — | 64 KB (per SM) |
Cache L2 | — | 3 MB |
Pixel rate | 8.808 GPixel/s | 79.20 GPixel/s |
---|---|---|
Texture rate | 35.23 GTexel/s | 224.4 GTexel/s |
FP16 (half) performance | 2.255 TFLOPS (2:1) | 14.36 TFLOPS (2:1) |
FP32 (float) performance | 1,127 GFLOPS | 7.181 TFLOPS |
FP64 (double) performance | 70.46 GFLOPS (1:16) | 224.4 GFLOPS (1:32) |
Slot width | IGP | Dual-slot |
---|---|---|
TDP | 35 W | 184 W |
Outputs | No outputs | 1x DVI1x HDMI2x DisplayPort1x USB Type-C |
Power connectors | None | 1x 8-pin |
Length | — | 229 mm 9 inches |
Width | — | 113 mm 4.4 inches |
Height | — | 35 mm 1.4 inches |
Suggested PSU | — | 450 W |
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 | — | Dec 7th, 2021 |
---|---|---|
Generation | — | GeForce 20 |
Predecessor | — | GeForce 10 |
Successor | — | GeForce 30 |
Production | — | Active |
Interface | — | PCIe 3.0 x16 |
Reviews | — | 86 in our database |