Cornelis and Synopsys Achieve Best-in-Class Ansys Fluent Performance Powered by AMD

Cornelis CN5000 with AMD EPYC 9755 processors beat previous record by >2X, boosting productivity and reducing time-to-market for engineering simulation and design teams

WAYNE, Pa.--(BUSINESS WIRE)--Cornelis, a leading provider of intelligent, high-performance networking solutions, today announced a new best-in-class performance score for Ansys Fluent, now part of Synopsys, achieved on the industry’s most demanding real-world computational fluid dynamics (CFD) benchmark. The result underscores the power of the Cornelis CN5000 Omni-Path interconnect paired with AMD EPYC™ processors to accelerate engineering simulation workloads at scale.

“This is a great example of the tangible benefits that are possible when the network stops being a bottleneck,” said Lisa Spelman, CEO, Cornelis. “Together with AMD and Synopsys, we’re empowering engineers in sectors like automotive, aerospace, and manufacturing to explore more ideas, test them faster, and turn bold concepts into real-world breakthroughs.”

Achieving this level of performance is a major productivity breakthrough for engineers who rely on simulation to design vehicles, aircraft, and advanced industrial systems. By sustaining nearly 99% strong-scaling efficiency on an eight-node AMD EPYC 9755 processor-based cluster, the combined solution doubles throughput, enabling more design iterations, higher-fidelity models, and faster time-to-insight within the same schedule and budget.

“Engineering simulation workloads demand both high-performance compute and networking efficiency,” said Robert Hormuth, Corporate Vice President, Architecture and Strategy, AMD. “AMD EPYC processors are designed to scale and deliver consistent, high-throughput performance for these types of applications. This result demonstrates how compute platforms built on EPYC can provide engineers with the speed and efficiency needed to accelerate innovation.”

The CN5000 end-to-end 400Gbps portfolio, comprising an ultra-low latency PCIe 5.0 SuperNIC and a 48-port, full non-blocking edge switch, has become the preferred HPC interconnect of choice for AMD EPYC processor-based clusters across high performance computing industries that demand both compute intensity and communication efficiency.

Why It Matters

In industries like automotive, aerospace, and manufacturing, where every design decision can impact safety, reliability, and millions in warranty or recall costs, simulation speed is more than a performance metric; it’s a strategic advantage. When simulations run faster, teams can complete more design studies in the same amount of time, accelerating innovation while reducing the risk of costly design flaws.

With CN5000 sustaining near-perfect scaling at eight AMD EPYC 9755 processor-based nodes with a total of over 1024 CPU cores, engineering teams can now:

• Explore more design variants per day, enabling broader evaluation of aerodynamic, thermal, and acoustic scenarios.
• Model higher-fidelity physics to capture complex, real-world behaviors like turbulence, vibration, and stress loads.
• Reduce time-to-market by compressing development cycles without compromising quality or safety.

The result: products that reach the market faster, perform better, and meet the highest standards for safety and reliability, proving that open, standards-based collaboration among hardware, networking, and software leaders delivers measurable impact where it matters most.

“The CN5000 accelerates simulation workloads at scale to shorten time-to-insight,” said Wim Slagter, Senior Director, Partner Programs at Synopsys. “This means engineers can run twice as many design studies in the same timeframe to expand design exploration, and make better decisions informed by high-fidelity physics. Together, we're enabling faster product development cycles while improving product functionality.”

A New Benchmark for Engineering Simulation

The Ansys Fluent f1_racecar_140m benchmark is one of the largest publicly available CFD test cases, widely used across the automotive and aerospace industries to evaluate simulation performance. The new score of 1704 achieved on 1024 AMD EPYC 9755 CPU cores with Zen 5 architecture and Cornelis CN5000 which is more than double the previous record of 830 achieved using the exact same number of prior-generation AMD EPYC 7773X CPUs and HDR 200G interconnects.

This record validates CN5000’s lossless, congestion-free architecture and AMD EPYC processors as a powerful combination for communication-heavy engineering simulation workloads. Engineers using Ansys Fluent 2025 R1 can expect near-linear scaling performance across multi-node systems, minimizing cost per iteration and maximizing engineering throughput.

For more on the CN5000 Omni-Path product family that’s helping organizations accelerate AI and HPC workloads, visit: https://www.cornelisnetworks.com/products/cn5000.

About Cornelis Networks

Cornelis Networks delivers high-performance, scale-out networking solutions that accelerate AI and HPC workloads. Built on the powerful Omni-Path architecture, Cornelis technology enables lossless, congestion-free networking that reduces training time, improves inference, and maximizes compute utilization. From foundation model training to complex climate modeling and real-time analytics, Cornelis' solutions power the most demanding workloads across commercial, academic, government and cloud environments. With a focus on performance, scalability, and efficiency, Cornelis helps organizations achieve faster insights and greater return on infrastructure investments. Learn more at cornelisnetworks.com.

AMD, the AMD Arrow logo and the combination thereof are trademarks of Advanced Micro Devices, Inc. Other names are for informational purposes only and may be trademarks of their respective owners.