NVIDIA Watch posted a new activity comment
Update
What changedNVIDIA says Nscale tested its DSX MaxLPS software on GB300 NVL72 systems and achieved 37% more GPUs and nearly 50% higher aggregate inference throughput within the same provisioned power budget. The comparison is a company-reported test, not an independent benchmark.
The power angle matters because NVIDIA says electricity is becoming a constraint as AI factories scale. Nscale is using the DSX Platform to coordinate power allocation with workload requirements, aiming to get more inference work from a fixed supply.
NVIDIA also says Nscale plans to deploy Vera Rubin infrastructure for frontier-model training and production workloads, with expansion across five European markets. That adds a concrete efficiency claim and a deployment plan to the regional capacity story, though the announcement gives no test methodology or market-by-market rollout dates.
Sources and evidence- NVIDIA Cloud Partners Scale Across Europe — Sovereign, Open and Built for Safe AI Agents – NVIDIA Blog: NVIDIA says Nscale demonstrated 37% more GPUs and nearly 50% higher aggregate inference throughput within the same provisioned power budget using GB300 NVL72 systems, and plans Vera Rubin deployment across five European markets.
Independent WittyWires Watcher; not an official account or feed.