by MicrosoftUS
Provide Azure customers with high-performance, power-efficient server CPUs for cloud workloads, reducing costs and improving AI workload throughput.
In early access preview; broader customer availability underway in 2026.
Updated Jun 2026Checked 11 Oct0 updates this week
Two-chiplet design improves manufacturability on N3 and enables flexible multi-socket configurations.
Each core runs independently at different voltages and frequencies, optimizing power for varied workloads.
3MB L2 per core plus 192MB L3 system cache reduce memory latency and off-chip bandwidth demands.
Dual-socket configurations support up to 264 total cores for distributed, parallel-heavy workloads.
| Spec | Azure Cobalt 200 |
|---|---|
| Active Cores | 132R (reported) |
| Core Architecture | ARM Neoverse V3R (reported) |
| Design | Two chiplets with 66 cores eachR (reported) |
| L2 Cache per Core | 3 MBR (reported) |
| L3 System Cache | 192 MBR (reported) |
| Manufacturing Process | TSMC N3R (reported) |
| vs. Cobalt 100 | +50%R (reported) |
| RAM Channels | 12R (reported) |
R reported by the company
Microsoft
AI models, cloud computing, Azure, GitHub Copilot
Microsoft develops software, cloud, and hardware across computing and AI. It operates Azure and GitHub. In June 2026, Microsoft launched frontier AI models in-house, gaining freedom after its OpenAI partnership ended in April 2026.
