● Covered by 1 source · 1 reportMedium impact1 neutral
Azure Cobalt 200 Improves EDA Workload Performance by up to 4.15x Over Cobalt 100
Key points
- Azure Cobalt 200 now hosts Microsoft's internal EDA workloads.
- Performance increased 1.30x to 4.15x over Cobalt 100.
- EDA tools from Cadence, Synopsys, and Siemens ran unmodified.
- Cobalt 200 offers flexibility for diverse EDA compute requirements.
Azure Cobalt 200 for Silicon Design
Microsoft has transitioned its internal electronic design automation (EDA) workloads to Azure Cobalt 200, a custom compute infrastructure built on the Arm Neoverse CSS V3. These EDA jobs are critical for designing, testing, and validating all Azure chips before manufacturing, encompassing millions of compute tasks annually.
The evaluation showed that EDA workloads, including simulation, verification, and static timing analysis, ran with 1.30x to 4.15x more performance on Cobalt 200 compared to Cobalt 100. This performance improvement was observed across various testing scenarios within Microsoft's silicon design environment.
Benefits of Cloud-Based EDA
Historically, EDA workloads relied on fixed, on-premise x86 server capacity, often leading to underutilization. The adoption of cloud-based EDA on Microsoft Azure provides flexibility, scalability, and reliability for compute-intensive silicon design. Using Azure Cobalt-based virtual machine instances, Microsoft's silicon engineering teams are achieving performance and efficiency gains in EDA workloads.
This approach optimizes the speed and performance of designing increasingly sophisticated silicon for complex cloud systems, a benefit also observed by Azure customers across other workloads.
Tool Compatibility and Workload Suitability
The evaluation included tools such as Cadence Xcelium, Cadence Spectre, Synopsys PrimeTime, Synopsys VCS, and Siemens Calibre. Tools that were enabled on Cobalt 100 functioned without modification on Cobalt 200, allowing the focus to remain on performance testing.
Cobalt 200 is suited for the diverse compute requirements of EDA, which include highly parallel circuit and analog simulation, throughput-driven functional and formal verification, and memory-intensive implementation and signoff tasks. Predictable per-core performance is important for planning overnight regressions in computer-aided design (CAD) management.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Microsoft has begun running its electronic design automation (EDA) workloads on Azure Cobalt 200, its custom Arm-based compute infrastructure. This shift resulted in performance gains of 1.30x to 4.15x compared to Cobalt 100 across various silicon design tasks. The move allows for greater flexibility and efficiency in designing complex silicon for cloud systems.