Can H100 80GB SXM5 run Phi-4?

NVIDIA · 80 GB HBM3 · 3350 GB/s bandwidth · Microsoft · 14.7B · model ctx up to 16K

Yes — the H100 80GB SXM5 (80 GB HBM3) can run Phi-4. At Q4 with the default 8K context it needs ≈ 12.03 GB VRAM (weights 8.09 GB + KV cache 2.94 GB + 1 GB runtime overhead), which fits within 80 GB. Decode at Q4/8K: ≈ 145.0 tok/s (estimated, batch 1). FP16 (≈ 33.34 GB) also fits, at ≈ 39.9 tok/s (estimated). 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	Yes	145.0 est.
Q4	8K default	12.03 GB	Yes	145.0 est.
Q4	32K	20.84 GB	Yes	145.0 est.
FP16	4K	31.87 GB	Yes	39.9 est.
FP16	8K default	33.34 GB	Yes	39.9 est.
FP16	32K	42.16 GB	Yes	39.9 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

Quant	Weights	KV cache	Overhead	Total needed	H100 80GB SXM5 VRAM
Q4	8.09 GB	2.94 GB	1 GB	12.03 GB	80 GB
FP16	29.4 GB	2.94 GB	1 GB	33.34 GB	80 GB

Other GPUs that run Phi-4

• 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.
• 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.
• RTX 4090D 24GB (China) — ≈43.6 tok/s Q4 est.

All GPUs for Phi-4 →

Other models for the H100 80GB SXM5

• Qwen3-0.6B — ≈2664.8 tok/s Q4 est.
• gemma-3-1b-it — ≈2131.8 tok/s Q4 est.
• Llama-3.2-1B-Instruct — ≈1776.5 tok/s Q4 est.
• Qwen3-1.7B — ≈1065.9 tok/s Q4 est.
• Qwen3-30B-A3B — ≈710.6 tok/s Q4 est.
• Qwen3-Coder-30B-A3B-Instruct — ≈710.6 tok/s Q4 est.
• GLM-4.7-Flash — ≈710.6 tok/s Q4 est.
• Qwen3-30B-A3B-Instruct-2507 — ≈710.6 tok/s Q4 est.

All models on the H100 80GB SXM5 →