Fully validated and packaged solution featuring ArcOS and Edgecore’s Tomahawk 5 platform delivers improved customer experience with disaggregated ultra-low latency, congestion-free networking for GPU clusters
Arrcus, the hyperscale networking software company and a leader in core, edge, and multi-cloud routing and switching infrastructure, today announced a collaboration with Edgecore Networks to deliver a turnkey solution optimized for AI/ML workloads and GPU-powered infrastructures. The integration of Arrcus’ ArcOS with Edgecore’s TH5 800G switch platform provides customers with a high-performance, congestion-free experience out-of-the box—enabling faster deployment, lower operational overhead, and superior workload efficiency.
As enterprises and cloud providers scale out GPU clusters to support AI training and inference, reliable high-throughput connectivity becomes critical. The combined Arrcus-Edgecore solution eliminates common bottlenecks by delivering:
- Out-of-the-box readiness with pre-validated hardware and software integration
- AI-optimized telemetry and traffic engineering for lossless Ethernet and RoCEv2 support
- Deterministic performance for RDMA workloads across large GPU fabrics
- Disaggregated design flexibility, enabling customers to scale across multiple form factors and deployment environments
“GPU clusters powering AI models are only as good as the network fabric that connects them,” said Shekar Ayyar, CEO of Arrcus. “Together with Edgecore, we’re simplifying that fabric, eliminating friction, and helping customers rapidly scale AI infrastructure with confidence.”
Edgecore’s AIS-800 series based on Broadcom’s Tomahawk 5 silicon, delivers industry-leading 51.2 Tbps switching capacity and native 800G ports in a compact 1RU form factor—making it an ideal match for data-intensive AI environments. Running ArcOS, the switch delivers precise traffic management, advanced ECN and PFC capabilities, and robust programmability to support dynamic workloads.
“This collaboration brings together Edgecore’s powerful, open hardware and Arrcus’ flexible software stack to deliver a complete plug-and-play networking solution for AI and data center operators,” said Mingshou Liu, President of Edgecore Networks. “This partnership reflects our ongoing commitment to the open infrastructure ecosystem, powered by decades of engineering expertise.”
This out-of-the-box, turnkey experience builds on Arrcus’ ongoing focus on simplifying network operations for cloud, edge, and core deployments—providing a single network operating system that runs across form factors, from Switches and Routers to VMs, Containers and SmartNICs.
About Arrcus
Arrcus is a leading provider of networking software solutions that empower businesses to achieve unparalleled scalability, performance, and reliability in their infrastructure. Arrcus is disrupting the industry with disaggregated solutions that deliver innovative, agile, and cost-effective networking, allowing enterprises to break free from traditional, monolithic systems and embrace a more flexible, efficient, and scalable approach to modern networking. The Arrcus team consists of world-class technologists who have an unparalleled record in shipping industry-leading networking products, complemented by industry thought leaders, operating executives, strategic partners, and top-tier VCs. The company is headquartered in San Jose, Calif. For more information, go to www.arrcus.com or follow Arrcus on LinkedIn and Twitter/X.
View source version on businesswire.com: https://www.businesswire.com/news/home/20250520156357/en/
Contacts
Media Contact
Sean Griffin
sean@arrcus.com
