.svg){kind=logo}
Discrete GPU
Intel Arc B580M
8 GB~7.2 GB usable
- Type
- Discrete GPU
- Provider
- Intel Arc
- VRAM
- 8 GB
- Price
- Unknown
Usable for models (est.): 7.2 GB · Price per GB: $0
Popular models that run here
The largest models this device holds fully in memory, each at its best-fitting quantization.
Phi-414B parametersBest fit: INT2
Llama 2 13B13B parametersBest fit: INT2
OLMo 2 13B13B parametersBest fit: INT2
Mistral Nemo 12B12B parametersBest fit: FP4
Gemma 3 12B12B parametersBest fit: FP4- Gemma 2 9B9B parametersBest fit: FP4
- Llama 3 8B8B parametersBest fit: FP4
- Llama 3.1 8B8B parametersBest fit: FP4
Every model: does it fit?
23 of 41 models run fully in memory, 7 more with partial offload. Estimated at a 4096-token context. How fit is calculated
| Model | Params | Best fit | Memory needed (est.) | Fit |
|---|---|---|---|---|
| DeepSeek-V4-Pro (MoE) | 1600B | — | ~480.0 GB | ✕ Does not fit |
| DeepSeek-V3 (MoE) | 671B | — | ~201.3 GB | ✕ Does not fit |
| Llama 3.1 405B | 405B | — | ~121.5 GB | ✕ Does not fit |
| DeepSeek-V4-Flash (MoE) | 284B | — | ~85.2 GB | ✕ Does not fit |
| Qwen 3 235B-A22B (MoE) | 235B | — | ~70.5 GB | ✕ Does not fit |
| Mistral Large 2 (MoE) | 141B | — | ~42.3 GB | ✕ Does not fit |
| Qwen 2.5 72B | 72B | — | ~21.6 GB | ✕ Does not fit |
| Llama 3 70B | 70.6B | — | ~21.2 GB | ✕ Does not fit |
| Llama 3.3 70B | 70.6B | — | ~21.2 GB | ✕ Does not fit |
| Llama 3.1 70B | 70B | — | ~21.0 GB | ✕ Does not fit |
| Llama 2 70B | 70B | — | ~21.0 GB | ✕ Does not fit |
| Falcon 40B | 40B | — | ~12.0 GB | ◐ Partial offload |
| Qwen 2.5 32B | 32B | — | ~9.6 GB | ◐ Partial offload |
| Qwen 3 32B | 32B | — | ~9.6 GB | ◐ Partial offload |
| OLMo 3 32B | 32B | — | ~9.6 GB | ◐ Partial offload |
| Gemma 4 31B | 31B | — | ~9.3 GB | ◐ Partial offload |
| Qwen 3 30B-A3B (MoE) | 30B | — | ~9.0 GB | ◐ Partial offload |
| Gemma 3 27B | 27B | — | ~8.1 GB | ◐ Partial offload |
| Phi-4 | 14B | INT2 | ~4.2 GB | ✓ Runs |
| Llama 2 13B | 13B | INT2 | ~3.9 GB | ✓ Runs |
| OLMo 2 13B | 13B | INT2 | ~3.9 GB | ✓ Runs |
| Mistral Nemo 12B | 12B | FP4 | ~7.2 GB | ✓ Runs |
| Gemma 3 12B | 12B | FP4 | ~7.2 GB | ✓ Runs |
| Gemma 2 9B | 9B | FP4 | ~5.4 GB | ✓ Runs |
| Llama 3 8B | 8B | FP4 | ~4.8 GB | ✓ Runs |
| Llama 3.1 8B | 8B | FP4 | ~4.8 GB | ✓ Runs |
| Qwen 3 8B | 8B | FP4 | ~4.8 GB | ✓ Runs |
| Llama 3.3 8B | 8B | FP4 | ~4.8 GB | ✓ Runs |
| Llama 2 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| Mistral 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| Qwen 2.5 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| Falcon 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| OLMo 2 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| OLMo-7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| Mistral Small 3 7B | 7B | FP4 | ~4.2 GB | ✓ Runs |
| Gemma 3 4B | 4B | FP8 | ~4.8 GB | ✓ Runs |
| Phi-3 | 3.8B | FP8 | ~4.6 GB | ✓ Runs |
| Phi-3-mini-128K | 3.8B | FP8 | ~4.6 GB | ✓ Runs |
| Llama 3.2 3B | 3B | FP16 | ~7.2 GB | ✓ Runs |
| Llama 3.2 1B | 1B | FP32 | ~4.8 GB | ✓ Runs |
| Gemma 3 1B | 1B | FP32 | ~4.8 GB | ✓ Runs |
Measured throughput
No one has submitted a measured benchmark for this device yet. The fit list above is an estimate.
Similar GPUs
- NVIDIA RTX 4090 (24 GB)
- NVIDIA RTX 6000 Ada (48 GB)
- NVIDIA A100 40GB PCIe (40 GB)
- NVIDIA A100 80GB PCIe (80 GB)
- NVIDIA H100 PCIe (80 GB)
- NVIDIA H100 SXM (80 GB)
- NVIDIA RTX 4080 Super (16 GB)
- NVIDIA RTX 4070 Ti Super (16 GB)
Alternatives with similar memory
- MacBook Air M3 13-inch (8 GB)
- Mac Mini M3 (8 GB)
- Qualcomm Snapdragon 7+ Gen 3 Leading Version (8 GB)
- Qualcomm Snapdragon 7 Gen 1 (8 GB)