GPU model | GP104 | TU104 |
---|---|---|
GPU variant | GP104-300-A1 | TU104-150-KC-A1 |
Architecture | Pascal | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 16 nm | 12 nm |
Number of transistors | 7,200 million | 13,600 million |
Die size | 314 mm² | 545 mm² |
Release date | Nov 2nd, 2017 | Jan 10th, 2020 |
---|---|---|
Generation | GeForce 10 | GeForce 20 |
Predecessor | GeForce 900 | GeForce 10 |
Successor | GeForce 20 | GeForce 30 |
Production | End-of-life | Active |
Launch price | 399 USD | 299 USD |
Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Reviews | 81 in our database | 85 in our database |
Base clock | 1607 MHz | 1365 MHz |
---|---|---|
Boost clock | 1683 MHz | 1680 MHz |
Memory clock | 2002 MHz 8 Gbps effective | 1750 MHz 14 Gbps effective |
Memory size | 8 GB | 6 GB |
---|---|---|
Memory type | GDDR5 | GDDR6 |
Memory bus width | 256 bit | 192 bit |
Bandwidth | 256.3 GB/s | 336.0 GB/s |
Shading units | 2432 | 1920 |
---|---|---|
TMUs | 152 | 120 |
ROPs | 64 | 48 |
SM count | 19 | 30 |
Cache L1 | 48 KB (per SM) | 64 KB (per SM) |
Cache L2 | 2 MB | 3 MB |
Tensor cores | — | 240 |
RT cores | — | 30 |
Pixel rate | 107.7 GPixel/s | 80.64 GPixel/s |
---|---|---|
Texture rate | 255.8 GTexel/s | 201.6 GTexel/s |
FP16 (half) performance | 127.9 GFLOPS (1:64) | 12.90 TFLOPS (2:1) |
FP32 (float) performance | 8.186 TFLOPS | 6.451 TFLOPS |
FP64 (double) performance | 255.8 GFLOPS (1:32) | 201.6 GFLOPS (1:32) |
Slot width | Dual-slot | Dual-slot |
---|---|---|
Length | 267 mm 10.5 inches | 229 mm 9 inches |
Width | 112 mm 4.4 inches | 113 mm 4.4 inches |
Height | 40 mm 1.6 inches | 35 mm 1.4 inches |
TDP | 180 W | 160 W |
Suggested PSU | 450 W | 450 W |
Outputs | 1x DVI1x HDMI3x DisplayPort | 1x DVI1x HDMI2x DisplayPort1x USB Type-C |
Power connectors | 1x 8-pin | 1x 8-pin |
Board number | PG411 | PG161 SKU 46 |
DirectX | 12 (12_1) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.2 | 1.2 |
CUDA | 6.1 | 7.5 |
Shader model | 6.4 | 6.6 |