GPU model | GP104 | TU116 |
---|---|---|
GPU variant | GP104-140-KA-A1 | TU116-250-KA-A1 |
Architecture | Pascal | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 16 nm | 12 nm |
Number of transistors | 7,200 million | 6,600 million |
Die size | 314 mm² | 284 mm² |
Release date | Dec 25th, 2016 | Nov 22nd, 2019 |
---|---|---|
Generation | GeForce 10 | GeForce 16 |
Predecessor | GeForce 900 | GeForce 10 |
Successor | GeForce 20 | GeForce 20 |
Production | End-of-life | Active |
Launch price | 199 USD | 159 USD |
Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Reviews | 117 in our database | 27 in our database |
Base clock | 1506 MHz | 1530 MHz |
---|---|---|
Boost clock | 1708 MHz | 1725 MHz |
Memory clock | 2002 MHz 8 Gbps effective | 1500 MHz 12 Gbps effective |
Memory size | 3 GB | 4 GB |
---|---|---|
Memory type | GDDR5 | GDDR6 |
Memory bus width | 192 bit | 128 bit |
Bandwidth | 192.2 GB/s | 192.0 GB/s |
Shading units | 1152 | 1280 |
---|---|---|
TMUs | 72 | 80 |
ROPs | 48 | 32 |
SM count | 9 | 20 |
Cache L1 | 48 KB (per SM) | 64 KB (per SM) |
Cache L2 | 1536 KB | 1024 KB |
Pixel rate | 81.98 GPixel/s | 55.20 GPixel/s |
---|---|---|
Texture rate | 123.0 GTexel/s | 138.0 GTexel/s |
FP16 (half) performance | 61.49 GFLOPS (1:64) | 8.832 TFLOPS (2:1) |
FP32 (float) performance | 3.935 TFLOPS | 4.416 TFLOPS |
FP64 (double) performance | 123.0 GFLOPS (1:32) | 138.0 GFLOPS (1:32) |
Slot width | Dual-slot | Dual-slot |
---|---|---|
Length | 250 mm 9.8 inches | 229 mm 9 inches |
TDP | 120 W | 100 W |
Suggested PSU | 300 W | 300 W |
Outputs | 1x DVI1x HDMI3x DisplayPort | 1x DVI1x HDMI1x DisplayPort |
Power connectors | 1x 6-pin | 1x 6-pin |
Width | — | 111 mm 4.4 inches |
Height | — | 35 mm 1.4 inches |
Board number | — | PG166 SKU 24 |
DirectX | 12 (12_1) | 12 (12_1) |
---|---|---|
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 |