GPU model | Polaris 21 | GF100 |
---|---|---|
GPU variant | Polaris 21M XT | GF100-850-A3 |
Architecture | GCN 4.0 | Fermi |
Foundry | GlobalFoundries | TSMC |
Manufacturing process | 14 nm | 40 nm |
Number of transistors | 3,000 million | 3,100 million |
Die size | 123 mm² | 529 mm² |
Release date | Apr 11th, 2018 | — |
---|---|---|
Generation | Mobility Radeon (RX M500X) | — |
Production | Active | — |
Interface | MXM-B (3.0) | — |
Predecessor | Crystal System | — |
GPU clock | 1275 MHz | 574 MHz |
---|---|---|
Memory clock | 1450 MHz 5.8 Gbps effective | 750 MHz 3 Gbps effective |
Shader clock | — | 1147 MHz |
Memory size | 4 GB | 3 GB |
---|---|---|
Memory type | GDDR5 | GDDR5 |
Memory bus width | 128 bit | 384 bit |
Bandwidth | 92.80 GB/s | 144.0 GB/s |
Shading units | 1024 | 448 |
---|---|---|
TMUs | 64 | 56 |
ROPs | 16 | 48 |
Compute units | 16 | — |
Cache L1 | 16 KB (per CU) | 64 KB (per SM) |
Cache L2 | 1024 KB | 768 KB |
SM count | — | 14 |
Pixel rate | 20.40 GPixel/s | 16.07 GPixel/s |
---|---|---|
Texture rate | 81.60 GTexel/s | 32.14 GTexel/s |
FP16 (half) performance | 2.611 TFLOPS (1:1) | — |
FP32 (float) performance | 2.611 TFLOPS | 1,028 GFLOPS |
FP64 (double) performance | 163.2 GFLOPS (1:16) | 513.9 GFLOPS (1:2) |
Slot width | MXM Module | Dual-slot |
---|---|---|
TDP | 65 W | 238 W |
Outputs | No outputs | 1x DVI |
Power connectors | None | 1x 6-pin + 1x 8-pin |
Length | — | 248 mm 9.8 inches |
Suggested PSU | — | 550 W |
DirectX | 12 (12_0) | 12 (11_0) |
---|---|---|
OpenGL | 4.6 | 4.6 |
OpenCL | 2.1 | 1.1 |
Vulkan | 1.2 | — |
Shader model | 6.4 | 5.1 |
CUDA | — | 2.0 |
Release date | — | Jul 25th, 2011 |
---|---|---|
Generation | — | Tesla |
Production | — | End-of-life |
Interface | — | PCIe 2.0 x16 |