Joint evaluation pairs the congestion-free CN5000 fabric with the Maverick-2 dataflow accelerator to attack the two bottlenecks that idle AI and HPC systems.

Summary:

• Cornelis and NextSilicon are collaborating to develop next-generation AI and HPC reference architectures.
• The effort combines congestion-free networking and reconfigurable compute to address infrastructure bottlenecks that leave expensive systems underutilized.
• Joint evaluations are designed to deliver validated system blueprints for OEM partners and customers.
• Future work will focus on emerging AI workloads, including disaggregated inference and agentic AI.

Cornelis and NextSilicon today announced at ISC High Performance 2026 a collaboration to build and evaluate joint reference architectures for AI and high-performance computing. The work pairs the Cornelis CN5000 fabric with the NextSilicon Maverick-2 compute platform. Joint evaluation is already underway, with the goal of commercialization through joint OEM partners.

The collaboration starts with the 400 Gbps CN5000 fabric, launched in 2025, paired with Maverick-2, which began shipping in volume late that year. The first phase validates how fabric and compute perform together across configurations, so OEM partners start from proven combinations rather than untested parts lists. The companies plan to extend testing to the 800 Gbps CN6000 fabric, due in the second half of 2026.

Two Bottlenecks, One Design

Each company targets a different bottleneck. Standard Ethernet was not built for the small, latency-sensitive messages that AI inference and HPC simulation generate at scale. Congestion builds, and expensive compute sits idle waiting on data. The CN5000 is designed to eliminate that idle time.

On the compute side, the von Neumann model that has defined processors for decades shuttles data between memory and a fixed execution unit. It stalls on the irregular, data-dependent workloads that now dominate AI and HPC. NextSilicon built Maverick-2 on its Intelligent Compute Architecture (ICA), a software-defined dataflow design that reconfigures to each workload at runtime and runs existing code without modification.

Pairing the two addresses both limits at once: a fabric that keeps data moving and an accelerator that keeps compute busy. The joint reference architectures will give OEM partners a blueprint for systems they can build and bring to market.

“Operators keep telling us their most expensive systems sit idle, waiting on the network," said Lisa Spelman, CEO of Cornelis. "We built the CN5000 to end that wait. NextSilicon challenges the same kind of assumption on the compute side, so this collaboration is a natural fit. Together we can show partners and customers what a congestion-free fabric and a workload-driven compute architecture deliver as one design.”

"For decades, software had to bend to fit the processor," said Elad Raz, founder and CEO of NextSilicon. "Maverick-2 makes the processor adapt to the software. Cornelis takes the same approach to the network. Evaluating our architectures together is the first step toward giving customers and OEM partners a faster, more efficient foundation for AI and HPC."

Looking Ahead: Disaggregated Inference

Along with HPC, the collaboration will also target the shift in AI inference toward Mixture of Experts (MoE) models and agentic AI. Production inference for these workloads no longer runs as one model on one accelerator. Inference splits into stages, and data moves between stages across the network.

This pattern, often called disaggregated inference, makes the fabric part of the compute path. It rewards a network that moves small, bursty, latency-sensitive messages without congestion, and compute that adapts to each stage of the pipeline. As the CN6000 reaches availability in the second half of 2026, the companies intend to evaluate how a congestion-free fabric and a reconfigurable compute architecture can support disaggregated and agentic inference, with findings intended to inform future OEM reference designs.

See us at ISC 2026

Visit Cornelis at booth E02 and NextSilicon in the virtual exhibitor’s hall at ISC High Performance 2026, June 23-25, at Congress Center Hamburg in Hamburg, Germany.

About Cornelis

Cornelis delivers high-performance scale-out and scale-up networking solutions that accelerate AI and HPC workloads. 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 www.cornelis.com.

About NextSilicon

NextSilicon builds computing infrastructure for algorithmically complex workloads. The company's Maverick-2 accelerator uses a runtime reconfigurable dataflow architecture to deliver up to 10x performance over leading GPUs at less than half the power, with no requirement to rewrite existing applications. Maverick-2 is in production at customer sites across HPC, AI, and national security computing environments. NextSilicon is headquartered in Tel Aviv, Israel, with offices in Minneapolis, MN, in the United States.

For more information, please visit www.nextsilicon.com

FAQ

1. What is Cornelis and NextSilicon announcing?
   Cornelis and NextSilicon are collaborating to build and evaluate joint reference architectures for AI and high-performance computing (HPC) that combine the CN5000 fabric with the Maverick-2 compute platform.
2. What problem does this collaboration address?
   The collaboration targets two common infrastructure bottlenecks: network congestion and compute inefficiency. The goal is to improve system utilization and application performance for AI and HPC workloads.
3. Who will benefit from this work?
   OEM partners and customers can benefit from validated reference architectures designed to simplify deployment and accelerate time to value.
4. What types of workloads are being evaluated?
   The companies are evaluating AI inference and HPC workloads today, with future plans to explore emerging use cases such as Mixture of Experts (MoE) models, agentic AI, and disaggregated inference.
5. What comes next?
   Joint evaluations are already underway, with commercialization expected through OEM partners. Future testing is planned on the 800 Gbps CN6000 fabric as it becomes available.

View source version on businesswire.com: https://www.businesswire.com/news/home/20260622687428/en/