The use case: how Databricks deploys CoreWeave
Databricks used NVIDIA H100 GPUs hosted by CoreWeave to complete the 8k-context training phase of its MPT-30B model, moving training over from A100 clusters mid-run. CoreWeave also supported the team as it got up to speed with the new GPU architecture.
More reported results
Figures published by Databricks on its own page. They are not independently measured by usedby.
In mid-June, after we received our first batch of 256 NVIDIA H100 GPUs from CoreWeave, we seamlessly moved MPT-30B to the new cluster to resume training on H100s with an average MFU of >35%.
Other AI tools used by Databricks
Other public relationships recorded on usedby.
Explore Databricks’s AI stack
Other companies using CoreWeave
Customers across industries, with public evidence.
General IntuitionConfirmed
IBMConfirmed
Mercado LibreConfirmed
MirisConfirmed
Mistral AIConfirmed
NovelAIConfirmedGo straight to the source
Based on Databricks’s own page about how it uses CoreWeave. We check the public evidence; the results are reported by the company.
databricks.com- Source published
- Jun 22, 2023
- Last checked
- Nov 9, 2026
Open source capture What the story claims, and what we checked
What we compared with the page.
- The figure: 2.44xChecked
Printed word for word on the page, near the name of Databricks.
- The passage quoted aboveChecked
Copied word for word from the page, near the name of Databricks.
- The numbers in our summaryChecked
Each one is printed on the page.
- The publication dateChecked
Read from the page’s own metadata, never guessed.
- Databricks uses CoreWeaveChecked
Confirmed line.
- The result itselfNot checked
We quote it; we did not measure it.
In mid-June, after we received our first batch of 256 NVIDIA H100 GPUs from CoreWeave, we seamlessly moved MPT-30B to the new cluster to resume training on H100s with an average MFU of >35%.