NVIDIA Deploys Vera CPU to Speed Up EDA Workflows for Its Next-Generation Chips

What Is the Vera CPU's Hardware Architecture?

According to NVIDIA's blog post, the Vera CPU combines 88 custom NVIDIA Olympus CPU cores with what NVIDIA describes as a "high-efficiency" LPDDR5X memory subsystem and a second-generation NVIDIA Scalable Coherent Fabric. NVIDIA states this combination is designed to deliver strong per-core performance, high memory bandwidth, and low latency for engineering applications.

NVIDIA's blog also includes an image caption confirming where this hardware currently runs: a "Vera cluster in NVIDIA Portland data center." That places the deployment NVIDIA is describing inside its own Portland, Oregon facility rather than a third-party or customer site.

Why Is NVIDIA Working With Cadence and Synopsys?

NVIDIA says it is collaborating with Cadence and Synopsys to optimize critical electronic design automation (EDA) applications specifically for the Vera CPU. The company frames the goal as using its own chip to speed up the process of designing chips: NVIDIA states it "is now deploying Vera across EDA workflows used to develop its next generation of CPUs and GPUs, demonstrating how high-performance CPU architecture can help accelerate some of the industry's most demanding engineering workloads."

In other words, by NVIDIA's own account, Vera is not just a product being sold — it is also infrastructure the company is running internally to build the chips that come after it.

Which EDA Tools Were Optimized, and What Do They Do?

Two specific applications are named in NVIDIA's post. Cadence Jasper is described as a formal verification platform that "uses smart proof technology and machine learning to find and fix bugs and improve verification productivity early in the design cycle." Synopsys VCS is described as a "high-performance functional verification solution used to simulate and validate complex chip designs before fabrication."

NVIDIA's post notes that in its test, Synopsys VCS "used the same number of cores" as the Jasper test — indicating the comparison between the two tools was run under matched core-count conditions on Vera.

What Performance Gains Did NVIDIA Report?

NVIDIA states that both applications — Cadence Jasper and Synopsys VCS — "showed up to 1.5x higher performance on selected workloads." This is the only performance figure NVIDIA discloses in the post, and it applies to both tools despite them coming from two different vendors and serving different functions (formal verification versus functional simulation).

What Comes After Vera in NVIDIA's CPU Roadmap?

NVIDIA says it plans to build on Vera with a next-generation CPU called Rosa, which it states will be "powered by the NVIDIA Rigel core." The company adds that it will keep "optimizing leading EDA applications across its CPU roadmap," indicating the same Cadence/Synopsys tuning work described for Vera is intended to continue for future generations.

What This Means

Taken together, NVIDIA's own disclosures describe a closed loop: the 88-core Vera CPU, running in NVIDIA's Portland data center, is being used inside the EDA workflows that NVIDIA relies on to design its next CPUs and GPUs — including Vera's own successor, Rosa, built on the Rigel core. The one quantified result NVIDIA provides, up to 1.5x higher performance on selected workloads, is attributed identically to both Cadence Jasper and Synopsys VCS, two tools from competing EDA vendors serving different verification functions. NVIDIA's post does not explain why both tools reportedly saw the same ceiling, and no independent benchmark is cited — the 1.5x figure comes solely from NVIDIA's own blog.