Chips · accelerators · NVIDIA
NVIDIA H200 SXM · Hopper
1,414 gigaflops of dense BF16 compute per watt, at 700 W
Shipped 28 Aug 2024; shipped after the first date promised, up to 2 months. Epoch AI names it at 3 AI data centers and 6 GPU clusters.
Embed
The frame credits NVIDIA, Epoch AI, AI data centers and Epoch AI, Data on GPU clusters and links back to its page. Preview it.
Cite
Each figure in NVIDIA’s own words
one chip, dense compute; memory is high-bandwidth memory- BF16, dense
- 990 TFLOPS
BFLOAT16 Tensor Core2 1,979 TFLOPS
NVIDIA H200 GPU (halved, because the page gives it with sparsity) - FP8, dense
- 1,979 TFLOPS
FP8 Tensor Core2 3,958 TFLOPS
NVIDIA H200 GPU (halved, because the page gives it with sparsity) - Memory
- 141 GB
GPU Memory 141GB 141GB
NVIDIA H200 GPU (read from a table with a column for each product, whose head reads “H200 SXM1 H200 NVL1”) - Memory bandwidth
- 4.8 TB/s
GPU Memory Bandwidth 4.8TB/s 4.8TB/s
NVIDIA H200 GPU (read from a table with a column for each product, whose head reads “H200 SXM1 H200 NVL1”) - Maximum power
- 700 W
Max Thermal Design Power (TDP) Up to 700W (configurable)
NVIDIA H200 GPU - Compute per watt
- 1,414 GFLOPS/Wdense BF16 over maximum power; 2,827 GFLOPS/W at FP8
The H200 page prints only sparse Tensor Core figures (footnote 2); dense is half. The delivered date is the earnings release that reports the first general availability, not the day it began.
1 dated promise and the shipment
each date links to the page it is quoted from- 13 Nov 2023 Promised H200 available from system makers and cloud providers starting in Q2 2024 in the second quarter of 2024.
The NVIDIA H200 will be available from global system manufacturers and cloud service providers starting in the second quarter of 2024.
NVIDIA Supercharges Hopper, the World's Leading AI Computing Platform - 28 Aug 2024 Shipped: H200 systems generally available on CoreWeave.
Revealed that H200 GPU-powered systems are now available on CoreWeave, the first cloud service provider to announce general availability.
NVIDIA Announces Financial Results for Second Quarter Fiscal 2025
Named at 3 AI data centers
H100-equivalents and power are the whole site’s today, with every chip it holds| Site | Owner | Country | H100e | MW | Chips |
|---|---|---|---|---|---|
| Colossus 1 | SpaceXAI | United States | 275,796 | 425 | H100, H200, B200 |
| AWS New Albany | Amazon | United States | 201,112 | 255 | Trainium2, H100, H200, B200 |
| Nebius Mantsala | Nebius | Finland | 48,528 | 90 | H100, B300, H200 |
126,752 chips in 6 GPU clusters
Epoch AI’s list of AI supercomputers, which stops in July 2025| Cluster | Owner | Country | Chips | First operational |
|---|---|---|---|---|
| Oracle OCI Supercluster H200s | Oracle | United States of America | 65,536 | 2024-11-21 |
| CoreWeave H200s | CoreWeave | United States of America | 42,000 | 2024-08-28 |
| AIST ABCI 3.0 | National Institute of Advanced Industrial Science and Technology (AIST) | Japan | 6,128 | 2025-01-20 |
| Nebius ISEG2 | Nebius AI | Iceland | 4,992 | 2025-06-15 |
| Sesterce Synapse Phase 2 | Sesterce | France | 4,096 | 2025-01-15 |
| Nebius Kansas City Phase 1 | Nebius AI | United States of America | 4,000 | 2025-04-15 |
Epoch AI’s names for this chip: H200; NVIDIA H200 SXM.
Other NVIDIA chips
- Tesla V100 SXM2 417 gigaflops of dense FP16 compute per watt, at 300 W
- A100 SXM 80GB 780 gigaflops of dense BF16 compute per watt, at 400 W
- H100 SXM 1,414 gigaflops of dense BF16 compute per watt, at 700 W
- GH200 Grace Hopper Superchip 990 gigaflops of dense BF16 compute per watt, at 1,000 W
- B200 2,200 gigaflops of dense BF16 compute per watt, at 1,000 W
- GB200 Grace Blackwell Superchip 5,000 teraflops of dense BF16 compute a chip; NVIDIA publishes no power figure
- B300 (Blackwell Ultra) 2,000 gigaflops of dense BF16 compute per watt, at 1,100 W
- Rubin GPU 4,000 teraflops of dense BF16 compute a chip; NVIDIA publishes no power figure
- NVIDIA’s own pages, each linked beside the figure or date it gives. Each page was read and each quote checked on 23 Oct 2026.
- Where it runs: Epoch AI, AI data centers and Epoch AI, GPU clusters, CC BY 4.0.