Can Arc B570 10GB run Phi-4-reasoning?

Intel · 10 GB GDDR6 · 380 GB/s bandwidth · Microsoft · 14.7B · model ctx up to 32K

No — Phi-4-reasoning at Q4 needs ≈ 10.56 GB even at 4K context, beyond the Arc B570 10GB's 10 GB GDDR6. It would only run with heavy CPU/disk offload. Decode at Q4/8K: ≈ 32.5 tok/s (estimated, batch 1). FP16 needs ≈ 33.34 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

Quant	Context	VRAM needed	Fits?	tok/s (decode)
Q4	4K	10.56 GB	No	42.1 est.
Q4	8K default	12.03 GB	No	32.5 est.
Q4	32K	20.84 GB	No	10.8 est.
FP16	4K	31.87 GB	No	1.3 est.
FP16	8K default	33.34 GB	No	1.2 est.
FP16	32K	42.16 GB	No	—

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

Quant	Weights	KV cache	Overhead	Total needed	Arc B570 10GB VRAM
Q4	8.09 GB	2.94 GB	1 GB	12.03 GB	10 GB
FP16	29.4 GB	2.94 GB	1 GB	33.34 GB	10 GB

Other GPUs that run Phi-4-reasoning

• B200 192GB SXM — ≈346.3 tok/s Q4 est.
• MI300X 192GB — ≈229.4 tok/s Q4 est.
• H200 141GB SXM5 — ≈207.8 tok/s Q4 est.
• H100 80GB SXM5 — ≈145.0 tok/s Q4 est.
• Gaudi 3 128GB — ≈141.8 tok/s Q4 est.
• A100 80GB SXM4 — ≈88.3 tok/s Q4 est.
• RTX 5090 32GB — ≈77.6 tok/s Q4 est.
• RTX 4090 24GB — ≈43.6 tok/s Q4 est.

All GPUs for Phi-4-reasoning →

Other models for the Arc B570 10GB

• Qwen3-0.6B — ≈302.3 tok/s Q4 est.
• gemma-3-1b-it — ≈241.8 tok/s Q4 est.
• Llama-3.2-1B-Instruct — ≈201.5 tok/s Q4 est.
• Qwen3-1.7B — ≈120.9 tok/s Q4 est.
• Llama-3.2-3B-Instruct — ≈75.6 tok/s Q4 est.
• Phi-4-mini-instruct — ≈63.6 tok/s Q4 est.
• Phi-4-mini-reasoning — ≈63.6 tok/s Q4 est.
• gemma-3-4b-it — ≈62.0 tok/s Q4 est.

All models on the Arc B570 10GB →