GPU model | TU106 | GA103 |
---|---|---|
GPU variant | TU106-125-A1 | GA103 |
Architecture | Turing | Ampere |
Foundry | TSMC | Samsung |
Manufacturing process | 12 nm | 8 nm |
Number of transistors | 10,800 million | unknown |
Die size | 445 mm² | unknown |
Release date | Jun 18th, 2020 | — |
---|---|---|
Generation | GeForce 16 | — |
Predecessor | GeForce 10 | — |
Successor | GeForce 20 | — |
Production | Active | — |
Interface | PCIe 3.0 x16 | — |
Base clock | 1410 MHz | 585 MHz |
---|---|---|
Boost clock | 1590 MHz | 1125 MHz |
Memory clock | 1500 MHz 12 Gbps effective | 1500 MHz 12 Gbps effective |
Memory size | 4 GB | 16 GB |
---|---|---|
Memory type | GDDR6 | GDDR6 |
Memory bus width | 128 bit | 256 bit |
Bandwidth | 192.0 GB/s | 384.0 GB/s |
Shading units | 896 | 7680 |
---|---|---|
TMUs | 56 | 240 |
ROPs | 32 | 96 |
SM count | 14 | 60 |
Tensor cores | 112 | 240 |
RT cores | 14 | 60 |
Cache L1 | 64 KB (per SM) | 128 KB (per SM) |
Cache L2 | 1024 KB | 4 MB |
Pixel rate | 50.88 GPixel/s | 108.0 GPixel/s |
---|---|---|
Texture rate | 89.04 GTexel/s | 270.0 GTexel/s |
FP16 (half) performance | 5.699 TFLOPS (2:1) | 17.28 TFLOPS (1:1) |
FP32 (float) performance | 2.849 TFLOPS | 17.28 TFLOPS |
FP64 (double) performance | 89.04 GFLOPS (1:32) | 270.0 GFLOPS (1:64) |
Slot width | Dual-slot | — |
---|---|---|
Length | 229 mm 9 inches | — |
Width | 111 mm 4.4 inches | — |
Height | 35 mm 1.4 inches | — |
TDP | 90 W | 95 W |
Suggested PSU | 250 W | — |
Outputs | 1x DVI1x HDMI1x DisplayPort | No outputs |
Power connectors | 1x 6-pin | None |
DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 7.5 | 8.6 |
Shader model | 6.6 | 6.5 |
Release date | — | 2022 |
---|---|---|
gpu.details.availability | — | — |
Generation | — | GeForce 30 Mobile |
Predecessor | — | GeForce 20 Mobile |
Production | — | Unreleased |
Interface | — | PCIe 4.0 x16 |