NVIDIA · Hopper

What runs on a H100 SXM 80GB

80 GB at 3350 GB/s. The largest model that fits at Q4_K_M is gpt-oss 120B-A5.1B, at roughly 323 tokens per second.

Model

Hardware

Quantisation

Context 8K

KV cache precision

Runtime

Batch

GPUs

Monitor plugged into this GPU

Runs comfortably

6.4 GB of 74.4 GB · 9%

074 GB

Weights 4.6 GB

KV cache 1.0 GB

Runtime overhead 0.8 GB

Llama 3.1 8B at Q4_K_M leaves 68.0 GB spare on a H100 SXM 80GB. There is room to raise the context length or move up a quantisation level.

Generation422tok/s

Prompt processing25864tok/s

Max context128Ktokens

KV per 1K tokens0GB

Every quantisation of Llama 3.1 8B on a H100 SXM 80GB

Highlighted row is the highest quality that still fits at 8K context.

Quantisation	bpw	Weights	Total	Verdict	Max ctx	tok/s
FP16 / BF16	16.00	15.0 GB	16.8 GB	Runs comfortably	128K	147
INT8 / W8A8	8.50	7.7 GB	9.5 GB	Runs comfortably	128K	271
Q8_0 (GGUF)	8.50	7.7 GB	9.5 GB	Runs comfortably	128K	271
FP8 (E4M3)	8.00	7.5 GB	9.3 GB	Runs comfortably	128K	278
Q6_K	6.56	6.1 GB	8.0 GB	Runs comfortably	128K	331
AWQ 4-bit	4.25	5.4 GB	7.2 GB	Runs comfortably	128K	369
GPTQ 4-bit	4.25	5.4 GB	7.2 GB	Runs comfortably	128K	369
MXFP4	4.25	5.4 GB	7.2 GB	Runs comfortably	128K	369
Q5_K_M	5.67	5.3 GB	7.1 GB	Runs comfortably	128K	375
Q5_K_S	5.52	5.2 GB	7.0 GB	Runs comfortably	128K	384
Q4_K_M	4.85	4.6 GB	6.4 GB	Runs comfortably	128K	422
Q4_K_S	4.58	4.4 GB	6.2 GB	Runs comfortably	128K	439
Q4_0	4.55	4.4 GB	6.2 GB	Runs comfortably	128K	441
IQ4_XS	4.25	4.1 GB	5.9 GB	Runs comfortably	128K	462
Q3_K_M	3.91	3.8 GB	5.7 GB	Runs comfortably	128K	489
IQ3_M	3.70	3.7 GB	5.5 GB	Runs comfortably	128K	507
IQ3_XXS	3.06	3.2 GB	5.0 GB	Runs comfortably	128K	571
Q2_K	2.63	2.8 GB	4.6 GB	Runs comfortably	128K	624
IQ2_XXS	2.06	2.3 GB	4.2 GB	Runs comfortably	128K	712
IQ1_M	1.75	2.1 GB	3.9 GB	Runs comfortably	128K	770

Q4_K_M at 8K context. Bandwidth sets the speed; capacity sets the ceiling.

Model	Params	Weights	Verdict	Max ctx	tok/s
DeepSeek-R1 671B-A37B	671B	379.0 GB	Won't fit	—	2.53
Qwen3 235B-A22B	235.1B	132.8 GB	Won't fit	—	6.80
gpt-oss 120B-A5.1B	116.8B	66.0 GB	Fits, but tight	108K	323
Llama 4 Scout 109B-A17B	109B	61.7 GB	Runs comfortably	63K	177
Qwen2.5 72B	72.7B	41.2 GB	Runs comfortably	103K	53.8
Llama 3.1 70B	70.5B	40.0 GB	Runs comfortably	107K	55.4
Mixtral 8x7B	46.7B	26.4 GB	Runs comfortably	32K	220
Command R 35B	35.0B	20.1 GB	Runs comfortably	43K	91.5
Yi-1.5 34B	34.4B	19.5 GB	Runs comfortably	32K	110
Qwen3 32B	32.8B	18.6 GB	Runs comfortably	128K	114
Qwen2.5-Coder 32B	32.8B	18.6 GB	Runs comfortably	128K	114
DeepSeek-R1-Distill-Qwen 32B	32.8B	18.6 GB	Runs comfortably	128K	114
Qwen3 30B-A3B	30.5B	17.3 GB	Runs comfortably	128K	192
Gemma 3 27B	27.4B	15.6 GB	Runs comfortably	128K	136
Mistral Small 24B	23.6B	13.4 GB	Runs comfortably	32K	161
gpt-oss 20B-A3.6B	20.9B	11.9 GB	Runs comfortably	128K	470
Qwen3 14B	14.8B	8.5 GB	Runs comfortably	128K	243
Phi-4 14B	14.7B	8.4 GB	Runs comfortably	16K	241
Mistral NeMo 12B	12.3B	7.0 GB	Runs comfortably	128K	286
Gemma 3 12B	12.2B	7.0 GB	Runs comfortably	128K	286
GLM-4 9B	9.4B	5.4 GB	Runs comfortably	128K	383
Qwen3 8B	8.2B	4.7 GB	Runs comfortably	128K	405
Llama 3.1 8B	8.0B	4.6 GB	Runs comfortably	128K	422
DeepSeek-R1-Distill-Llama 8B	8.0B	4.6 GB	Runs comfortably	128K	422
Qwen2.5 7B	7.6B	4.4 GB	Runs comfortably	128K	469
Mistral 7B v0.3	7.3B	4.1 GB	Runs comfortably	32K	463
Gemma 3 4B	4.3B	2.5 GB	Runs comfortably	128K	742
Qwen3 4B	4.0B	2.3 GB	Runs comfortably	128K	697
Llama 3.2 3B	3.2B	1.8 GB	Runs comfortably	128K	877
Llama 3.2 1B	1.2B	0.7 GB	Runs comfortably	128K	2192