Can GR00T N1.7 run on H100 80GB?

43.67 Hz inference on NVIDIA H100 80GB (tensorrt, bf16, 2 cam). Measured inference cost only, not task success. Sources: https://caniembody.pages.dev/compute/h100-80gb

Can MolmoAct2 run on H100 80GB?

runs on NVIDIA H100 80GB (pytorch-eager, bf16, 2 cam). Measured inference cost only, not task success. Sources: https://caniembody.pages.dev/compute/h100-80gb

Matrix/Compute/Datacenter

What can I run on H100 80GB?

80 GB VRAMHopperFP8 Tensor Core* 3,958 teraFLOPS (H100 SXM; *with sparsity)Up to 700W (configurable) (SXM)View matrix for this target →

The SXM part is 80GB HBM3; the H100 NVL is a different 94GB part. Not listed as a Hugging Face Jobs flavor in the current pricing table.

Policy by policy

Policy	Params	Inference memory (sourced, with precision)	On H100 80GB	Best measured
GR00T N1.7 VLA	3B	≥16 GB	measured	43.67 Hz · TensorRT · bf16 · 2 cam
MolmoAct2 VLA	—	12.1 GB measured bf16	measured	runs · PyTorch eager · bf16 · 2 cam
ACT Imitation	80M	211 MB measured fp32	no data
Diffusion Policy Imitation	—	1.12 GB measured fp32	no data
EO-1 VLA	3B	≥6.5 GB	no data
Evo-1 VLA	770M	—	no data
Fast-WAM World model	6B	—	no data
HIL-SERL RL	—	—	no data
LingBot-VA World model	5.3B	≥18 GB	no data
Multitask DiT Imitation	450M	—	no data
OpenVLA VLA	7.5B	15 GB measured bf16	no data
OpenVLA-OFT VLA	—	≥16 GB	no data
SmolVLA VLA	450M	1.75 GB measured fp32	no data
TD-MPC RL	—	—	no data
VLA-JEPA VLA	—	—	no data
VQ-BeT Imitation	—	—	no data
WALL-OSS VLA	—	—	no data
X-VLA VLA	900M	—	no data
π0 VLA	3.3B	13.32 GB measured fp32	no data
π0-FAST VLA	—	≥8 GB	no data
π0.5 VLA	—	≥8 GB	no data

“Exceeds memory” compares a policy's upstream-stated minimum inference memory (never a measured peak, which is precision-specific) with this target's VRAM. It is arithmetic, not observation, and is labelled inferred. Quantised or split runtimes can change the answer; a measured run always wins.

All measured runs on H100 80GB

Policy	Runtime	Precision	Cams	p50 / p95 ms	Inference	Mem	Platform	Source
GR00T N1.7	TensorRT	bf16	2 @256x256	22.9 / —	43.67 Hz	—	— conditions torch_threads: 8 intra-op / 1 inter-op Inference cost only; not task success. Median of 60 timed calls (3 repetitions x 20 after 10 warmup), fixed inputs, includes preprocessing and action decoding, excludes RPC. Precision is the repo default (unquantized BF16).	T3 · community reported
GR00T N1.7	torch.compile	bf16	2 @256x256	53.4 / —	18.73 Hz	—	— conditions torch_threads: 8 intra-op / 1 inter-op Inference cost only; not task success. Median of 60 timed calls (3 repetitions x 20 after 10 warmup), fixed inputs, includes preprocessing and action decoding, excludes RPC. Precision is the repo default (unquantized BF16).	T3 · community reported
GR00T N1.7	PyTorch eager	bf16	2 @256x256	101.5 / —	9.85 Hz	—	— conditions torch_threads: 8 intra-op / 1 inter-op Inference cost only; not task success. Median of 60 timed calls (3 repetitions x 20 after 10 warmup), fixed inputs, includes preprocessing and action decoding, excludes RPC. Precision is the repo default (unquantized BF16).	T3 · community reported
MolmoAct2	PyTorch eager	bf16	2	—	—	13 GB	lerobot 0.6.1 conditions os: Ubuntu 22.04 (repo tested with) · memory_sampling: nvidia-smi peak Inference memory only (12.1 GiB peak GPU memory = 13.0 GB); no latency reported. LIBERO with two RGB cameras, chunk_size=10, 8 flow timesteps. Not task success.	T3 · community reported

Sources

T1 · upstream documentedvendor-docsverified 2026-10-24 (1d ago)
H100 product page spec table (columns: H100 SXM, H100 NVL): FP8 3,958 TFLOPS sparse; memory 80GB vs 94GB.
“FP8 Tensor Core * 3,958 teraFLOPS 3,341 teraFLOPS INT8 Tensor Core * 3,958 TOPS 3,341 TOPS GPU Memory 80GB 94GB”
https://www.nvidia.com/en-us/data-center/h100/
T1 · upstream documentedvendor-docsverified 2026-10-24 (1d ago)
Same table: Max TDP up to 700W (SXM) and 350-400W (NVL).
“Max Thermal Design Power (TDP) Up to 700W (configurable) 350-400W (configurable)”
https://www.nvidia.com/en-us/data-center/h100/