Modal Selects Nvidia Vera CPUs After Tests Show Double Throughput Over Intel and AMD Chips

AI cloud provider Modal has selected Nvidia’s Vera CPUs for its infrastructure, citing internal tests that showed the new processors delivering twice the throughput of existing Intel and AMD chips. According to Tom’s Hardware, which reported the design win, Modal believes the performance gains justify the engineering effort required to adopt the new hardware for its agentic AI services.

Modal currently relies exclusively on AMD and Intel processors to support AI developers deploying agents that perform complex, autonomous operations over extended periods. These workloads demand substantial CPU resources for software execution and data processing tasks that GPUs cannot handle efficiently. Preliminary evaluations by Modal indicated that Nvidia’s Vera CPUs could significantly improve performance in these specific scenarios, leading to the strategic partnership.

Architectural Trade-offs and Performance Claims

The reported performance advantage is notable because agentic AI workloads often involve numerous independent tasks competing for CPU cores, cache capacity, and memory bandwidth. However, Modal has not disclosed the specific benchmark configurations, the exact AMD or Intel processor models used for comparison, or performance-per-watt figures. Consequently, it remains unclear whether Vera’s advantage stems from its Arm v9.2-based Olympus CPU cores, its memory subsystem offering 1.2 TB/s bandwidth, or other architectural features.

Nvidia’s Vera CPUs are based on the Arm instruction set architecture, whereas Modal’s current infrastructure runs on x86 processors from Intel and AMD. This fundamental difference means Modal must make substantial modifications to its software to support the new hardware. Despite the need for significant engineering changes, the company stated that the potential throughput improvements make the transition worthwhile for its specific use cases.

Market Impact and Supplier Strategy

Although Modal has secured this major design win with Nvidia, the company does not intend to cease using CPUs from its current suppliers. Reuters reports that Modal will continue to utilize Intel and AMD processors alongside the new Nvidia hardware. This approach suggests a hybrid infrastructure strategy rather than a complete replacement of existing x86-based systems with Arm-based alternatives.

The comparison between Vera and current Intel and AMD processors does not account for upcoming generations of x86 chips that will compete against Vera in the coming years. Tom’s Hardware notes that the tests compared Vera against currently available processors, leaving the long-term competitive landscape between Nvidia’s Arm-based offerings and next-generation x86 silicon undefined.

What should you do?

Security and technology professionals should monitor how Modal integrates Arm-based processors into its cloud infrastructure and whether this shift affects software compatibility for existing clients. Organizations deploying agentic AI workloads should evaluate their own CPU resource requirements and consider testing alternative architectures to determine if similar performance gains are achievable. Infrastructure teams should prepare for potential software modifications needed to support Arm-based workloads, as the industry moves toward heterogeneous CPU environments.

What to do and how to stay safe: Nvidia

• Evaluate current CPU utilization for agentic AI workloads to identify bottlenecks that may benefit from new architectures.
• Review software compatibility requirements when considering a shift from x86 to Arm-based processors in cloud environments.
• Monitor vendor announcements for updated benchmark data comparing current and next-generation CPU offerings.
• Assess the total cost of engineering effort required to port applications to new instruction set architectures.

Step-by-step guide: Mobile Device Security: How to Protect Data Beyond Passwords

General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

Will Modal stop using Intel and AMD processors?

No, Modal does not intend to cease using CPUs from its current suppliers and will continue to rely on AMD and Intel processors alongside Nvidia Vera.

How much faster is Nvidia’s Vera CPU compared to Intel and AMD?

Modal’s preliminary evaluations showed that Vera could deliver twice the throughput of its existing AMD and Intel processors in specific AI agent workloads.

What architectural change does Modal need to make for this upgrade?

Because Vera is based on the Arm instruction set architecture, Modal must make substantial modifications to its software to support the new processors.