Independent testing confirms DriveNets Network Cloud-AI is highest-performance Ethernet solution for AI
DriveNets – a leader in innovative networking solutions – today announced it has joined theUltra Ethernet Consortium, an organization aimed at building an open, interoperable, Ethernet-based architecture for high-performance networking serving AI and HPC workloads. The inclusion of DriveNets in the consortium reflects the company’s role as a leader and innovator in AI networking solutions. DriveNets Network Cloud-AI – introduced in May and now available for order with Broadcom Jericho 3-AI based white boxes – offers the highest-performance AI networking solution available today, based on trials with leading hyperscalers and independent testing byScala Computing. It improves the performance of large-scale AI workloads regardless of GPU or AI accelerator used, enabling users to build open, multi-vendor AI infrastructures.
DriveNets is part of the first group of companies to join the UEC after its founding by members including Microsoft, Meta, Broadcom, AMD, Arista, Cisco, Oracle, HPE, Intel and Eviden.
“AI is going to be ubiquitous in cloud and data center infrastructure and will require technology that can effectively minimize GPU idle cycles in high-performance AI workloads,” said Ido Susan, founder and CEO of DriveNets. “Solving for that calls for Ethernet-based fabric that supports open, standards-based architectures that eliminate vendor lock-in. The Ultra Ethernet Consortium is committed to advancing open, interoperable Ethernet-based solutions, and DriveNets is proud to support its efforts to evolve the Ethernet standard.”
DriveNets Network Cloud-AI solution is already 18-24 months ahead of the rest of the AI networking industry in terms of performance thanks to its implementation of high-scale, cell-based fabric with cell-based load balancing and end-to-end scheduling. Based on the highest-scale router technology, it uniquely supports AI back-end and front-end networks with the same solution.
The results of Scala Computing’s testing showed that DriveNets Network Cloud-AI Ethernet solution delivers exceptional job completion time (JCT) results at the highest workload performance without being locked to a specific hardware – supporting an open and flexible ecosystem of ASICs, ODMs, and optical vendors.
DriveNets also continues to contribute to the Open Compute Project (OCP) distributed disaggregated chassis (DDC) standard, which is becoming the leading open networking standard. OCP is also working with the UEC to improve Ethernet performance for AI clusters and HPC. Joining the UEC and Network Cloud-AI’s industry-leading Ethernet-based AI networking performance reinforce DriveNets’ leadership role in the efforts to bring high-performance Ethernet solutions to the industry.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!