- Graphics Processor: Raven
- Cores: 512
- TMUs / ROPs: 32 / 8
A budget AMD Radeon RX Vega 8 integrated graphics with limited performance overall, excellent energy efficiency. Suited for casual gaming and basic graphics tasks.
Review
Pros
- Great energy efficiency
- Dedicated AI tensor cores
- No discrete card needed — integrated solution
Cons
- Limited gaming performance at higher resolutions
- No hardware ray tracing
- Older generation architecture
Compare Radeon RX Vega 8 with other products
Full Specifications
| 1080p High | 1080p Ultra | 1440p Ultra | 4K Ultra | |
|---|---|---|---|---|
| Forza Horizon 5 | 13 | — | — | — |
| The Witcher 3 | 10 | — | — | — |
| Counter-Strike 2 | 15 | — | — | — |
| Far Cry 6 | 16 | — | — | — |
| Hogwarts Legacy | — | — | — | — |
| Call of Duty | 16 | — | — | — |
| Ghost of Tsushima | 9 | — | — | — |
| Cyberpunk 2077 | 9 | — | — | — |
| Tomb Raider | 16 | — | — | — |
| Average FPS | 12 | — | — | — |
Multiplatform graphics benchmark suite that directly correlates with performance in modern games
| Steel Nomad Lite Score | 1260 |
|---|---|
| Time Spy | 1410 |
| Solar Bay | — |
| Port Royal | — |
| Fire Strike | 4499 |
| Wild Life Extreme | 2532 |
| Night Raid | 17993 |
GPU test for computational tasks (image processing, photography, computer vision, and ML)
| GB6 Compute Score | 10472 |
|---|---|
| Background Blur | 21.2 img/sec |
| Face Detection | 10.9 img/sec |
| Horizon Detection | 0.34 Gpixels/sec |
| Edge Detection | 0.65 Gpixels/sec |
| Gaussian Blur | 0.51 Gpixels/sec |
| Feature Matching | 0.13 Gpixels/sec |
| Stereo Matching | 34.9 Gpixels/sec |
| Particle Physics | 1322.9 FPS |
| API | OpenCL |
Videocard test that focuses on compute shaders, multi-texturing, tessellation, and other features
Rendering performance test for 3D modeling
Tests throughput of AI operations in single, half, and quantized precision
| GB6 ML Single Precision | 2257 |
|---|---|
| GB6 ML Half Precision | 2817 |
| GB6 ML Quantized | 1643 |
| Image Classification (SP) | 924 |
| Image Segmentation (HP) | 1148 |
| Image Super Resolution (Q) | 2211 |
| Face Detection (HP) | 2819 |
| Pose Estimation (Q) | 7106 |
| Text Classification (SP) | 832 |
| Machine Translation (HP) | 1021 |
| Object Detection (SP) | 1040 |
| Depth Estimation (Q) | 3885 |
| Style Transfer (SP) | 14495 |
| Framework | ONNX |
| Backend | DirectML |
| Vendor | Amd |
|---|---|
| Build | Integrated |
| Released | 13-Feb-18 |
| Case | Laptop |
| Purpose | Generic |
| Segment | Entry-level |
| Architecture | GCN 5.0 |
| GPU Codename | Raven |
| Recommended CPU | -AMD Ryzen 7 4800Uor above |
| Base Clock | 300 MHz |
|---|---|
| Boost Clock | 1100-2100 MHz (configurable) |
| Shading Units | 512 |
| Texture Mapping Units (TMUs) | 32 |
| Render Output Units (ROPs) | 8 |
| Compute Units (Pipelines) | 8 |
| Tensor Cores | No |
| Ray-tracing Cores | No |
| L1 Cache | — |
| L2 Cache | — |
| Instructions Per Cycle | 2 IPC |
| Pixel Fill Rate | 9-17 GPixel/s |
|---|---|
| Texture Fill Rate | 35-67 GTexel/s |
| FLOPS (FP32) | 1.1-2.2 TFLOPS |
| Interface | IGP |
|---|---|
| TGP | 15 W |
| Manufacturing | TSMC |
| Fabrication Process | 7 nm |
| Die Size | 210 mm² |
| Transistor Count | 4.94 billion |
| Transistor Density | 23.52 MTr/mm² |
| Max. Temperature | — |
| Memory Type | System Shared |
|---|
| DirectX | 12 |
|---|---|
| Vulkan | 1.3 |
| OpenGL | 4.6 |
| OpenCL | 2 |
| CUDA | — |
| Ray Tracing | No |
| DLSS | No |
| DisplayPort | — |
