GPU model | GK104 | TU102 |
---|---|---|
GPU variant | N15E-GT-A2 | TU102-875-A1 |
Architecture | Kepler | Turing |
Foundry | TSMC | TSMC |
Manufacturing process | 28 nm | 12 nm |
Number of transistors | 3,540 million | 18,600 million |
Die size | 294 mm² | 754 mm² |
Release date | Mar 12th, 2014 | — |
---|---|---|
Generation | GeForce 800M | — |
Production | End-of-life | — |
Interface | MXM-B (3.0) | — |
Predecessor | GeForce 700M | — |
Successor | GeForce 900M | — |
Base clock | 941 MHz | 1065 MHz |
---|---|---|
Boost clock | 967 MHz | 1395 MHz |
Memory clock | 1250 MHz 5 Gbps effective | 1750 MHz 14 Gbps effective |
Memory size | 3 GB | 8 GB |
---|---|---|
Memory type | GDDR5 | GDDR6 |
Memory bus width | 192 bit | 384 bit |
Bandwidth | 120.0 GB/s | 672.0 GB/s |
Shading units | 1344 | 4608 |
---|---|---|
TMUs | 112 | 288 |
ROPs | 24 | 96 |
SMX count | 7 | — |
Cache L1 | 16 KB (per SMX) | 64 KB (per SM) |
Cache L2 | 384 KB | 6 MB |
SM count | — | 72 |
Tensor cores | — | 576 |
RT cores | — | 72 |
Pixel rate | 27.08 GPixel/s | 133.9 GPixel/s |
---|---|---|
Texture rate | 108.3 GTexel/s | 401.8 GTexel/s |
FP32 (float) performance | 2.599 TFLOPS | 12.86 TFLOPS |
FP64 (double) performance | 108.3 GFLOPS (1:24) | 401.8 GFLOPS (1:32) |
FP16 (half) performance | — | 25.71 TFLOPS (2:1) |
Slot width | MXM Module | Dual-slot |
---|---|---|
TDP | 100 W | 260 W |
Outputs | No outputs | No outputs |
Power connectors | None | 1x 6-pin + 1x 8-pin |
Board number | P2053 SKU 3 | PG150 SKU 220 |
Length | — | 267 mm 10.5 inches |
Suggested PSU | — | 600 W |
DirectX | 12 (11_0) | 12 Ultimate (12_2) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 3.0 | 3.0 |
Vulkan | 1.1 | 1.2 |
CUDA | 3.0 | 7.5 |
Shader model | 5.1 | 6.6 |
Release date | — | Unknown |
---|---|---|
Generation | — | GRID |
Production | — | Active |
Interface | — | PCIe 3.0 x16 |