GPU model | GA104 | GA102 |
---|---|---|
GPU variant | GA104-300-A1 | GA102-220-A1 |
Architecture | Ampere | Ampere |
Foundry | Samsung | Samsung |
Manufacturing process | 8 nm | 8 nm |
Number of transistors | 17,400 million | 28,300 million |
Die size | 392 mm² | 628 mm² |
Release date | Sep 1st, 2020 | Jan 11th, 2022 |
---|---|---|
gpu.details.availability | Oct 29th, 2020 | Jan, 2022 |
Generation | GeForce 30 | GeForce 30 |
Predecessor | GeForce 20 | GeForce 20 |
Production | Active | Active |
Launch price | 499 USD | 799 USD |
Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
Reviews | 78 in our database | — |
Base clock | 1500 MHz | 1260 MHz |
---|---|---|
Boost clock | 1725 MHz | 1710 MHz |
Memory clock | 1750 MHz 14 Gbps effective | 1188 MHz 19 Gbps effective |
Memory size | 8 GB | 12 GB |
---|---|---|
Memory type | GDDR6 | GDDR6X |
Memory bus width | 256 bit | 384 bit |
Bandwidth | 448.0 GB/s | 912.4 GB/s |
Shading units | 5888 | 8960 |
---|---|---|
TMUs | 184 | 280 |
ROPs | 96 | 112 |
SM count | 46 | 70 |
Tensor cores | 184 | 280 |
RT cores | 46 | 70 |
Cache L1 | 128 KB (per SM) | 128 KB (per SM) |
Cache L2 | 4 MB | 5 MB |
Pixel rate | 165.6 GPixel/s | 191.5 GPixel/s |
---|---|---|
Texture rate | 317.4 GTexel/s | 478.8 GTexel/s |
FP16 (half) performance | 20.31 TFLOPS (1:1) | 30.64 TFLOPS (1:1) |
FP32 (float) performance | 20.31 TFLOPS | 30.64 TFLOPS |
FP64 (double) performance | 317.4 GFLOPS (1:64) | 478.8 GFLOPS (1:64) |
Slot width | Dual-slot | Dual-slot |
---|---|---|
Length | 242 mm 9.5 inches | 285 mm 11.2 inches |
Width | 112 mm 4.4 inches | 112 mm 4.4 inches |
TDP | 220 W | 350 W |
Suggested PSU | 550 W | 750 W |
Outputs | 1x HDMI3x DisplayPort | 1x HDMI3x DisplayPort |
Power connectors | 1x 12-pin | 1x 12-pin |
Board number | PG142 SKU 10 | — |
Height | — | 40 mm 1.6 inches |
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 | 8.6 | 8.6 |
Shader model | 6.6 | 6.6 |