what/llm/can/i/run
gpus/nvidia/h100-pcie

H100 PCIe

Released 2022-03-22 · Street $28,000
vram
80 GB
discrete
bandwidth
2,000
GB/s
models fit
18
top · FP16
power
350 W
high

Spec sheet · top pick

Hardware spec
H100 PCIe
NVIDIA · Hopper · sm_90
vram
80 GB
bandwidth
2,000 GB/s
tdp
350 W
released
2022-03-22
msrp
$30,000
street
~$28,000
★ Sweet spot for this card
gemma 3 27b
FP16 · GGUF · 57.2 GB
bench50.6 · conf A
vram57.2 / 80 GB
speed35 tok/s
Best benchmark per GB of VRAM in your tier. FP16wastes <3% quality vs FP16, leaves headroom for 32k context.
Full model inspect →

Top 10 models on this GPU

#modelquantvramtok/sbench ↓fit
01gemma 3 27bFP1657.2 GB3550.6Great
02Qwen2.5 Coder 32BFP1667.7 GB3048.8Great
03Gemma 2 27BFP1657.2 GB3546.2Great
04Qwen3 32BFP1667.7 GB3040.0Great
05deepseek r1 distill llama 70bQ8_074 GB2751.7Ok
06Gemma 2 9BFP1619.4 GB10335.0Great
07Gemma 3 12BFP1625.7 GB7831.3Great
08deepseek r1 distill qwen 32bFP1667.7 GB3033.3Great
09Mixtral 8x22b v0.1Q4_K_M74.5 GB2720.0Ok
10Gemma 3 4BFP168.9 GB22515.1Great
Showing 10 of 18 models that fit · see full ranked list →

Upgrade options

★ Best value
+H200 SXM5
141 GB · ~$45,000 · +$17,000 over H100 PCIe
Unlocks +1 models not in current rank
See H200 SXM5
Bigger tier
+M1 Ultra 128GB
128 GB · ~$4,000 · +$-24,000 over H100 PCIe
Unlocks +1 models not in current rank
See M1 Ultra 128GB
Premium tier
+M2 Max 96GB
96 GB · ~$3,500 · +$-24,500 over H100 PCIe
Same model set · higher headroom for context / batch
See M2 Max 96GB

Also compare

H100 SXM5
80 GB · NVIDIA · 43% more
A100 80GB
80 GB · NVIDIA · 61% cheaper
M1 Max 64GB
64 GB · Apple unified · 92% cheaper
M1 Ultra 64GB
64 GB · Apple unified · 91% cheaper
Not running a H100 PCIe?
Tell us what you actually have — we'll re-rank everything for your machine.
Change my machine