GPU model | Navi 23 | TU106 |
---|---|---|
Architecture | RDNA 2.0 | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 7 nm | 12 nm |
Number of transistors | 11,060 million | 10,800 million |
Die size | 237 mm² | 445 mm² |
GPU variant | — | TU106-400A-A1 |
Release date | Jun 8th, 2021 | Oct 17th, 2018 |
---|---|---|
gpu.details.availability | Jul, 2021 | — |
Generation | Radeon Pro | GeForce 20 |
Production | Active | Active |
Launch price | 649 USD | 499 USD |
Interface | PCIe 4.0 x16 | PCIe 3.0 x16 |
Predecessor | — | GeForce 10 |
Reviews | — | 73 in our database |
Successor | — | GeForce 30 |
Base clock | 2331 MHz | 1410 MHz |
---|---|---|
Boost clock | 2903 MHz | 1620 MHz |
Memory clock | 1750 MHz 14 Gbps effective | 1750 MHz 14 Gbps effective |
Memory size | 8 GB | 8 GB |
---|---|---|
Memory type | GDDR6 | GDDR6 |
Memory bus width | 128 bit | 256 bit |
Bandwidth | 224.0 GB/s | 448.0 GB/s |
Shading units | 1792 | 2304 |
---|---|---|
TMUs | 112 | 144 |
ROPs | 64 | 64 |
Compute units | 28 | — |
RT cores | 28 | 36 |
gpu.details.l0-cache | 32 KB per WGP | — |
Cache L1 | 128 KB per Array | 64 KB (per SM) |
Cache L2 | 2 MB | 4 MB |
gpu.details.l3-cache | 32 MB | — |
SM count | — | 36 |
Tensor cores | — | 288 |
Pixel rate | 185.8 GPixel/s | 103.7 GPixel/s |
---|---|---|
Texture rate | 325.1 GTexel/s | 233.3 GTexel/s |
FP16 (half) performance | 20.81 TFLOPS (2:1) | 14.93 TFLOPS (2:1) |
FP32 (float) performance | 10.40 TFLOPS | 7.465 TFLOPS |
FP64 (double) performance | 650.3 GFLOPS (1:16) | 233.3 GFLOPS (1:32) |
Slot width | Single-slot | Dual-slot |
---|---|---|
Length | 241 mm 9.5 inches | 229 mm 9 inches |
TDP | 100 W | 175 W |
Suggested PSU | 300 W | 450 W |
Outputs | 4x DisplayPort | 1x DVI1x HDMI2x DisplayPort1x USB Type-C |
Power connectors | 1x 6-pin | 1x 8-pin |
Width | — | 113 mm 4.4 inches |
Height | — | 35 mm 1.4 inches |
Board number | — | PG160 SKU 52 |
DirectX | 12.0 Ultimate (12_2) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 3.0 |
Vulkan | 1.2 | 1.2 |
Shader model | 6.5 | 6.6 |
CUDA | — | 7.5 |