Intel Arc B580 12GBBudget pickC51.3 tok/s decode
~$249 MSRP
Can it run?
Tight fit
Using Q4_K_M in Ollama
Fit status
Tight fit
Decode
75.5 tok/s
TTFT
2563 ms
Safe context
17K
Memory
7.4 GB / 8.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs with offload (needs ~0.2 GB host RAM) | 70.1 tok/s | 4019 ms | 30K |
| Chat | C | Tight fit | 73.4 tok/s | 1438 ms | 9K |
| Coding | C | Tight fit | 75.5 tok/s | 2563 ms | 17K |
| RAG | C | Runs with offload (needs ~0.2 GB host RAM) | 70.1 tok/s | 5024 ms | 30K |
| Reasoning | C | Tight fit | 73.4 tok/s | 3115 ms |
How Llama 2 7B Chat (7B params) fits at each quantization level on RTX 3070 8GB (8.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 2.7 GB | Low | D37 |
Q3_K_S | 3 | 3.4 GB | Low | D39 |
NVFP4 | 4 |
Upgrade options
Intel Arc B580 12GBBudget pick~$249 MSRP
RTX 3060 12GBBest value~$329 MSRP
RTX 3080 10GBNVIDIA upgrade~$699 MSRP
RTX 2080 Ti 11GBBiggest leap~$999 MSRP
| 17K |
| Medium |
| D40 |
Q4_K_M | 4 | 4.3 GB | Medium | C41 |
Q5_K_MBest for your GPU | 5 | 5.0 GB | High | C43 |
Q6_K | 6 | 5.7 GB | High | C44 |
Q8_0 | 8 | 7.5 GB | Very High | C44 |
F16 | 16 | 14.3 GB | Maximum | F0 |