← AI Hashrate

Can Arc B580 12GB run DeepSeek-R1-Distill-Qwen-14B?

Intel · 12 GB GDDR6 · 456 GB/s bandwidth · DeepSeek · 14.8B · model ctx up to 128K

Partly — DeepSeek-R1-Distill-Qwen-14B at Q4 fits the Arc B580 12GB (12 GB GDDR6) at 4K context (≈ 10.62 GB) but not at the default 8K (≈ 12.1 GB needed). Use a shorter context or partial offload. Decode at Q4/8K: ≈ 55.1 tok/s (estimated, batch 1). FP16 needs ≈ 33.56 GB — does not fit on this card. Estimates come from memory-bandwidth math; rows tagged measured override estimates. Relative ranking is more reliable than absolute tok/s. Methodology.

Fit & speed by quant and context

QuantContextVRAM neededFits?tok/s (decode)
Q44K10.62 GBYes19.6 est.
Q48K default12.1 GBNo55.1 est.
Q432K20.98 GBNo18.3 est.
FP164K32.08 GBNo2.2 est.
FP168K default33.56 GBNo2.0 est.
FP1632K42.44 GBNo1.2 est.

Fits = weights + KV(ctx) + 1 GB overhead ≤ 95% of VRAM. Measured anchors are context-agnostic; the fit verdict is recomputed per context. A missing tok/s means the model is far beyond this card (offload-only territory).

VRAM breakdown at 8K context

QuantWeightsKV cacheOverheadTotal neededArc B580 12GB VRAM
Q48.14 GB2.96 GB1 GB12.1 GB12 GB
FP1629.6 GB2.96 GB1 GB33.56 GB12 GB

Other GPUs that run DeepSeek-R1-Distill-Qwen-14B

All GPUs for DeepSeek-R1-Distill-Qwen-14B →

Other models for the Arc B580 12GB

All models on the Arc B580 12GB →