The Extended Brief

NVIDIA Exemplar Cloud: Lessons for Unlocking Full Performance on AI Infrastructure

Brief by The AI News AI newsroom · Jul 30, 2026, 5:12 PM EDT edition

Original reporting by NVIDIA Developer Blog — Elizabeth Goodman · published Jul 30, 2026, 12:00 PM EDT

Updated Aug 1, 2026, 11:28 AM EDT

Misconfiguring kernel settings on new Blackwell or H100 clusters can silently waste up to 12% of your compute throughput, costing millions at scale.

Key points

• Identical NVIDIA H100, GB200 NVL72, or GB300 NVL72 clusters can yield materially different training throughput. source ↗
• Partner deployments show 8% to 12% throughput gaps compared to NVIDIA reference architectures. source ↗
• These performance gaps occur across identical workloads, models, and global batch sizes. source ↗
• Kernel configuration choices frequently cause these performance discrepancies between computing clusters. source ↗

Practical applications

• Audit kernel configuration on your H100, GB200 NVL72, or GB300 NVL72 clusters against NVIDIA's reference architectures before assuming hardware is the bottleneck.
• Run identical training workloads across your clusters and investigate any throughput gap in the 8-12 percent range as a likely configuration issue.
• Fold a configuration-parity check into cluster acceptance testing so new deployments are validated against reference throughput, not just uptime.

Context

Large AI training runs are executed on clusters of GPU systems such as NVIDIA's H100 and the newer GB200/GB300 NVL72 rack-scale designs, where small per-node inefficiencies compound across the whole fleet. NVIDIA's Exemplar Cloud findings show that two clusters built from identical hardware, running identical workloads and batch sizes, can differ materially in training throughput. The culprit is frequently kernel configuration choices rather than hardware, with partner deployments showing 8-12 percent gaps versus reference architectures.

What to watch

• Whether NVIDIA publishes concrete reference configurations or tooling that lets operators diagnose these kernel-level gaps automatically.
• Reports from other cluster operators confirming or narrowing the 8-12 percent throughput spread on identical hardware.

Related briefs

• Long Live the Short King: Why 4-hi HBM Wins
• Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra
• DeepSeek launching v4.1 flash cheaper and more capable than v4 pro
• AI coding startup Cognition raises $2B at $48B valuation as revenue nears $900M

This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.