Can Gaudi 3 128GB run NVIDIA-Nemotron-3-Nano-4B-BF16?

Intel · 128 GB HBM2e · 3276 GB/s bandwidth · NVIDIA · 4.0B · model ctx up to 256K

Yes — the Gaudi 3 128GB (128 GB HBM2e) can run NVIDIA-Nemotron-3-Nano-4B-BF16. At Q4 with the default 8K context it needs ≈ 4.0 GB VRAM (weights 2.2 GB + KV cache 0.8 GB + 1 GB runtime overhead), which fits within 128 GB. Decode at Q4/8K: ≈ 521.2 tok/s (estimated, batch 1). FP16 (≈ 9.8 GB) also fits, at ≈ 143.3 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	3.6 GB	Yes	521.2 est.
Q4	8K default	4.0 GB	Yes	521.2 est.
Q4	32K	6.4 GB	Yes	521.2 est.
FP16	4K	9.4 GB	Yes	143.3 est.
FP16	8K default	9.8 GB	Yes	143.3 est.
FP16	32K	12.2 GB	Yes	143.3 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	Gaudi 3 128GB VRAM
Q4	2.2 GB	0.8 GB	1 GB	4.0 GB	128 GB
FP16	8.0 GB	0.8 GB	1 GB	9.8 GB	128 GB

Other GPUs that run NVIDIA-Nemotron-3-Nano-4B-BF16

• B200 192GB SXM — ≈1272.7 tok/s Q4 est.
• MI300X 192GB — ≈843.2 tok/s Q4 est.
• H200 141GB SXM5 — ≈763.6 tok/s Q4 est.
• H100 80GB SXM5 — ≈533.0 tok/s Q4 est.
• A100 80GB SXM4 — ≈324.4 tok/s Q4 est.
• RTX 5090 32GB — ≈285.1 tok/s Q4 est.
• RTX 4090 24GB — ≈160.4 tok/s Q4 est.
• RTX 4090D 24GB (China) — ≈160.4 tok/s Q4 est.

All GPUs for NVIDIA-Nemotron-3-Nano-4B-BF16 →

Other models for the Gaudi 3 128GB

• MiniCPM5-1B — ≈1930.3 tok/s Q4 est.
• Qwen3.5-2B — ≈1042.4 tok/s Q4 est.
• GLM-4.7-Flash — ≈694.9 tok/s Q4 est.
• Qwen3-30B-A3B-Instruct-2507 — ≈694.9 tok/s Q4 est.
• Qwen3-Next-80B-A3B-Instruct — ≈694.9 tok/s Q4 est.
• Qwen3-Coder-Next — ≈694.9 tok/s Q4 est.
• Qwen3.5-35B-A3B — ≈694.9 tok/s Q4 est.
• Qwen3.6-35B-A3B — ≈694.9 tok/s Q4 est.

All models on the Gaudi 3 128GB →