AMD Radeon Pro WX Vega M GL vs AMD Radeon Vega 3 Embedded
Graphics Processor
| GPU model | Polaris 22 | Picasso |
|---|---|---|
| GPU variant | VegaM MGL XL | — |
| Architecture | GCN 4.0 | GCN 5.0 |
| Foundry | GlobalFoundries | GlobalFoundries |
| Manufacturing process | 14 nm | 14 nm |
| Number of transistors | 5,000 million | 4,940 million |
| Die size | 208 mm² | 210 mm² |
Mobile graphics
| Release date | Apr 24th, 2018 | — |
|---|---|---|
| Generation | Vega (Vega M) | — |
| Production | Active | — |
| Interface | IGP | — |
Clocks
| Base clock | 931 MHz | 300 MHz |
|---|---|---|
| Boost clock | 1011 MHz | 1000 MHz |
| Memory clock | 700 MHz 1400 Mbps effective | System Shared |
Memory Configuration
| Memory size | 4 GB | System Shared |
|---|---|---|
| Memory type | HBM2 | System Shared |
| Memory bus width | 1024 bit | System Shared |
| Bandwidth | 179.2 GB/s | System Dependent |
Render Configuration
| Shading units | 1280 | 192 |
|---|---|---|
| TMUs | 80 | 12 |
| ROPs | 32 | 4 |
| Compute units | 20 | 3 |
| Cache L1 | 16 KB (per CU) | — |
| Cache L2 | 1024 KB | — |
Performance
| Pixel rate | 32.35 GPixel/s | 4.000 GPixel/s |
|---|---|---|
| Texture rate | 80.88 GTexel/s | 12.00 GTexel/s |
| FP16 (half) performance | 2.588 TFLOPS (1:1) | 768.0 GFLOPS (2:1) |
| FP32 (float) performance | 2.588 TFLOPS | 384.0 GFLOPS |
| FP64 (double) performance | 161.8 GFLOPS (1:16) | 24.00 GFLOPS (1:16) |
Dimensions & Outputs
| Slot width | IGP | IGP |
|---|---|---|
| TDP | 65 W | 10 W |
| Outputs | No outputs | No outputs |
| Board number | D136 | — |
| Power connectors | — | None |
API Support & Features
| DirectX | 12 (12_0) | 12 (12_1) |
|---|---|---|
| OpenGL | 4.6 | 4.6 |
| OpenCL | 2.1 | 2.1 |
| Vulkan | 1.2 | 1.2 |
| Shader model | 6.4 | 6.4 |
Other Features
Integrated Graphics
| Release date | — | 2019 |
|---|---|---|
| Generation | — | Picasso (Vega) |
| Production | — | Active |
| Interface | — | IGP |
| Predecessor | — | Raven Ridge |
| Successor | — | Renoir |
