# H100 NVL

> NVIDIA · Hopper

Canonical: https://gpuninja.com/chips/h100-nvl
Data as of: 2026-10-11T10:37:40Z
Path: Home › GPU chips
- Specs verified: between 11 Oct 2026, 08:51 UTC and 11 Oct 2026, 10:37 UTC

## Specifications

- Vendor: NVIDIA (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Architecture: Hopper (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Die: GH100 (Verified · source pci-ids.ucw.cz · retrieved 11 Oct 2026, 08:51 UTC)
- Process: TSMC 4N (Verified · source nvidianews.nvidia.com · retrieved 11 Oct 2026, 07:13 UTC · checked 11 Oct 2026, 08:51 UTC)
- Shader cores: 14,592 (Verified · source lenovopress.lenovo.com · retrieved 11 Oct 2026, 07:03 UTC · checked 11 Oct 2026, 08:51 UTC)
- Tensor cores: 456 (Verified · source lenovopress.lenovo.com · retrieved 11 Oct 2026, 07:03 UTC · checked 11 Oct 2026, 08:51 UTC)
- RT cores: None
- Memory bus: 6,016-bit (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Max VRAM: 94 GB (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Reference boost clock: 1,785 MHz (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Reference TDP: 400 W (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Host interface: PCIe 5.0 x16 (Verified · source nvidia.com · retrieved 11 Oct 2026, 09:58 UTC · checked 11 Oct 2026, 10:37 UTC)
- Launch MSRP: No public MSRP
- Launch date: 21 Mar 2023 (Verified · source nvidianews.nvidia.com · retrieved 11 Oct 2026, 07:13 UTC · checked 11 Oct 2026, 08:51 UTC)
- Segment: Data center (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- PCI IDs: 10de:2321 (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:04 UTC · checked 11 Oct 2026, 08:51 UTC)
- Status: Active (Verified · source nvidia.com · retrieved 11 Oct 2026, 07:03 UTC · checked 11 Oct 2026, 08:51 UTC)

## Performance

- Memory bandwidth: 3,938 GB/s (Verified · source nvidia.com · retrieved 11 Oct 2026, 09:58 UTC · checked 11 Oct 2026, 10:37 UTC)
- FP32 compute: 60 TFLOPS (Verified · source nvidia.com · retrieved 11 Oct 2026, 09:58 UTC · checked 11 Oct 2026, 10:37 UTC)
- FP16 compute: 835.5 TFLOPS (Verified · source thundercompute.com · retrieved 11 Oct 2026, 10:00 UTC · checked 11 Oct 2026, 10:37 UTC)
- BF16 compute: 835.5 TFLOPS (Verified · source thundercompute.com · retrieved 11 Oct 2026, 10:00 UTC · checked 11 Oct 2026, 10:37 UTC)
- FP8 compute: 1,670.5 TFLOPS (Verified · source thundercompute.com · retrieved 11 Oct 2026, 10:00 UTC · checked 11 Oct 2026, 10:37 UTC)
- FP4 compute: Not supported
- INT8 compute: 1,670.5 TOPS (Verified · source thundercompute.com · retrieved 11 Oct 2026, 10:00 UTC · checked 11 Oct 2026, 10:37 UTC)

## Boards built on this chip

Board, Chip, VRAM, TDP, Length, Slots, Price, Stock

No boards recorded for this chip.

## Benchmarks

No benchmark results recorded.

## Sources

- https://www.nvidia.com/content/dam/en-zz/Solutions/Data-Center/h100/PB-11773-001_v01.pdf — Vendor, Architecture, Memory bus, Max VRAM, Reference boost clock, Reference TDP, Host interface, Segment, PCI IDs, Memory bandwidth; last verified 11 Oct 2026, 10:37 UTC
- https://pci-ids.ucw.cz/read/PC/10de/2321 — Die; last verified 11 Oct 2026, 08:51 UTC
- https://nvidianews.nvidia.com/news/nvidia-announces-hopper-architecture-the-next-generation-of-accelerated-computing — Process; last verified 11 Oct 2026, 08:51 UTC
- https://lenovopress.lenovo.com/lp1732-thinksystem-nvidia-h100-pcie-gen5-gpu — Shader cores, Tensor cores; last verified 11 Oct 2026, 08:51 UTC
- https://nvidianews.nvidia.com/news/nvidia-launches-inference-platforms-for-large-language-models-and-generative-ai-workloads — Launch date; last verified 11 Oct 2026, 08:51 UTC
- https://www.nvidia.com/en-us/data-center/h100/ — Status, FP32 compute; last verified 11 Oct 2026, 10:37 UTC
- https://www.thundercompute.com/blog/nvidia-h100-specs-full-guide — FP16 compute, BF16 compute, FP8 compute, INT8 compute; last verified 11 Oct 2026, 10:37 UTC