GPU model | TU104 | TU102 |
---|---|---|
GPU variant | TU104-400A-A1 | TU102-875-A1 |
Architecture | Turing | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 12 nm | 12 nm |
Number of transistors | 13,600 million | 18,600 million |
Die size | 545 mm² | 754 mm² |
Release date | Sep 20th, 2018 | Aug 13th, 2018 |
---|---|---|
Generation | GeForce 20 | Quadro |
Predecessor | GeForce 10 | — |
Production | Active | Active |
Launch price | 699 USD | 9,999 USD |
Current price | Amazon / Newegg | — |
Interface | PCIe 3.0 x16 | PCIe 3.0 x16 |
Reviews | 66 in our database | — |
Successor | GeForce 30 | — |
Base clock | 1515 MHz | 1230 MHz |
---|---|---|
Boost clock | 1710 MHz | 1620 MHz |
Memory clock | 1750 MHz 14 Gbps effective | 1750 MHz 14 Gbps effective |
Memory size | 8 GB | 48 GB |
---|---|---|
Memory type | GDDR6 | GDDR6 |
Memory bus width | 256 bit | 384 bit |
Bandwidth | 448.0 GB/s | 672.0 GB/s |
Shading units | 2944 | 4608 |
---|---|---|
TMUs | 184 | 288 |
ROPs | 64 | 96 |
SM count | 46 | 72 |
Tensor cores | 368 | 576 |
RT cores | 46 | 72 |
Cache L1 | 64 KB (per SM) | 64 KB (per SM) |
Cache L2 | 4 MB | 6 MB |
Pixel rate | 109.4 GPixel/s | 155.5 GPixel/s |
---|---|---|
Texture rate | 314.6 GTexel/s | 466.6 GTexel/s |
FP16 (half) performance | 20.14 TFLOPS (2:1) | 29.86 TFLOPS (2:1) |
FP32 (float) performance | 10.07 TFLOPS | 14.93 TFLOPS |
FP64 (double) performance | 314.6 GFLOPS (1:32) | 466.6 GFLOPS (1:32) |
Slot width | Dual-slot | Dual-slot |
---|---|---|
Length | 267 mm 10.5 inches | 267 mm 10.5 inches |
Width | 116 mm 4.6 inches | 111 mm 4.4 inches |
Height | 35 mm 1.4 inches | — |
TDP | 215 W | 260 W |
Suggested PSU | 550 W | 600 W |
Outputs | 1x HDMI3x DisplayPort1x USB Type-C | No outputs |
Power connectors | 1x 6-pin + 1x 8-pin | 1x 6-pin + 1x 8-pin |
Board number | PG180 SKU 2 | — |
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 | 7.5 |
Shader model | 6.6 | 6.6 |