GPU model | Mach64 | TU116 |
---|---|---|
GPU variant | Mach64 VT2 (215VT2CA42) | TU116-300-A1 |
Architecture | Mach | Turing |
Manufacturing process | 600 nm | 12 nm |
Number of transistors | 1 million | 6,600 million |
Die size | 90 mm² | 284 mm² |
Foundry | — | TSMC |
Release date | May 21st, 1996 | Mar 14th, 2019 |
---|---|---|
Generation | Mach 64 | GeForce 16 |
Production | End-of-life | Active |
Interface | PCI | PCIe 3.0 x16 |
Predecessor | Mach 32 | GeForce 10 |
Successor | Rage | GeForce 20 |
Launch price | — | 219 USD |
Reviews | — | 40 in our database |
GPU clock | 62 MHz | — |
---|---|---|
Memory clock | 62 MHz | 2001 MHz 8 Gbps effective |
Base clock | — | 1530 MHz |
Boost clock | — | 1785 MHz |
Memory size | 1 MB | 6 GB |
---|---|---|
Memory type | EDO | GDDR5 |
Memory bus width | 64 bit | 192 bit |
Bandwidth | 496.0 MB/s | 192.1 GB/s |
Pixel shaders | 1 | — |
---|---|---|
Vertex shaders | 0 | — |
TMUs | 0 | 88 |
ROPs | 1 | 48 |
Shading units | — | 1408 |
SM count | — | 22 |
Cache L1 | — | 64 KB (per SM) |
Cache L2 | — | 1536 KB |
Pixel rate | 62.00 MPixel/s | 85.68 GPixel/s |
---|---|---|
Texture rate | 0 MTexel/s | 157.1 GTexel/s |
FP16 (half) performance | — | 10.05 TFLOPS (2:1) |
FP32 (float) performance | — | 5.027 TFLOPS |
FP64 (double) performance | — | 157.1 GFLOPS (1:32) |
Slot width | Single-slot | Dual-slot |
---|---|---|
TDP | unknown | 120 W |
Suggested PSU | 200 W | 300 W |
Outputs | 1x VGA | 1x DVI1x HDMI1x DisplayPort |
Board number | 340 | PG165 SKU 0 |
Length | — | 229 mm 9 inches |
Width | — | 111 mm 4.4 inches |
Height | — | 35 mm 1.4 inches |
Power connectors | — | 1x 8-pin |
DirectX | — | 12 (12_1) |
---|---|---|
OpenGL | — | 4.6 |
OpenCL | — | 3.0 |
Vulkan | — | 1.2 |
Pixel shader | — | — |
Vertex shader | — | — |
CUDA | — | 7.5 |
Shader model | — | 6.6 |