Intel Arc A580 - 8GB usable
The entry point to 8GB Arc, fine for 7B experiments. Slowest of the line on the least-supported stack: buy it to learn, not to rely on.
Specifications
| Brand | Intel |
|---|---|
| Model | Arc A580 |
| Usable VRAM | 8GB |
| Architecture | Xe-HPG |
| CUDA / Stream Processors | 3,072 |
| Memory Bandwidth | 512 GB/s |
| TDP | 185W |
| FP32 TFLOPS | 10.4 |
Current Offers
Used from $215New from $165
Prices last updated:
GPUDojo is reader-supported. When you buy through links on our site, we may earn an affiliate commission.
Price History
- eBay$215mid-range
- Amazon$184current
- Newegg$165at high
For AI / LLM Use
Limited VRAM restricts you to 7B quantized models. Expect a smaller selection of supported local-AI builds than on NVIDIA.
What Models Can It Run?
- 7B Q6_K, 14B Q3_K (tight)
- 7B Q4_K_M only
Estimated Performance
Generation: ~38 tokens/sec
Prefill: ~186 tokens/sec
Recommended Quantisations
- Q4_K_M for 7B models
- Q3_K for larger experiments
Pros & Cons
Pros
- Consumer card: easy to install, display output
Cons
- Only 8GB usable VRAM: limited to small models
- Moderate memory bandwidth: not the fastest for inference
- Smaller local-AI software ecosystem than NVIDIA or AMD
Community Verdict
No community reviews yet for the Arc A580. Know a good review? Let us know.