GPU model | GA102 | GA104 |
---|---|---|
GPU variant | GA102-300-A1 | — |
Architecture | Ampere | Ampere |
Foundry | TSMC | Samsung |
Manufacturing process | 7 nm | 8 nm |
Number of transistors | 40,000 million | 17,400 million |
Die size | 627 mm² | 392 mm² |
Release date | 2020 | — |
---|---|---|
Generation | GeForce 30 | — |
Production | Unreleased | — |
Launch price | 1,399 USD | — |
Interface | PCIe 4.0 x16 | — |
Base clock | 1410 MHz | 1260 MHz |
---|---|---|
Boost clock | 1740 MHz | 1560 MHz |
Memory clock | 2375 MHz 19000 MHz effective | 1375 MHz 11 Gbps effective |
Memory size | 12 GB | 6 GB |
---|---|---|
Memory type | GDDR6X | GDDR6 |
Memory bus width | 384 bit | 192 bit |
Bandwidth | 912.0 GB/s | 264.0 GB/s |
Shading units | 5120 | 4096 |
---|---|---|
TMUs | 320 | 128 |
ROPs | 96 | 64 |
SM count | 80 | 32 |
Tensor cores | 640 | 128 |
RT cores | 160 | 32 |
Cache L1 | 192 KB (per SM) | 128 KB (per SM) |
Cache L2 | 6 MB | 4 MB |
Pixel rate | 167.0 GPixel/s | 99.84 GPixel/s |
---|---|---|
Texture rate | 556.8 GTexel/s | 199.7 GTexel/s |
FP16 (half) performance | 35.64 TFLOPS (2:1) | 12.78 TFLOPS (1:1) |
FP32 (float) performance | 17.82 TFLOPS | 12.78 TFLOPS |
FP64 (double) performance | 556.8 GFLOPS (1:32) | 199.7 GFLOPS (1:64) |
Slot width | Dual-slot | — |
---|---|---|
Length | 267 mm 10.5 inches | — |
TDP | 320 W | 130 W |
Suggested PSU | 700 W | — |
Outputs | 1x HDMI3x DisplayPort1x USB Type-C | No outputs |
Power connectors | 1x 12-pin | None |
Board number | PG132 SKU 20 | — |
DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.0 | 3.0 |
Vulkan | 1.2.140 | 1.2 |
CUDA | 8.5 | 8.6 |
Shader model | 6.5 | 6.6 |
Release date | — | Unknown |
---|---|---|
Generation | — | Quadro Mobile (Ax000) |
Production | — | Active |
Interface | — | PCIe 4.0 x16 |