AI Summit NY: DriveNets to Showcase Highest-Performance AI Networking Solutions

DriveNets, provider in high-scale networking solutions, will participate in AI Summit NY, taking place December 10-11, 2025 in New York City, NY.The company will demonstrate Network Cloud-AI, high-performance Ethernet-based AI networking solution supporting InfiniBand-level performance and used by hyperscalers, neoclouds, and enterprises.

Stop by Booth 317 to meet DriveNets’ AI networking experts and our infrastructure services team to explore our network fabric solutions for scale-up, scale-out, and scale-across.

Company’s successful deployment experiences and how solution delivers:

• Lossless standard Ethernet connectivity and the highest-performance multi-tenancy
• Highest-performance scale-out and scale-across multi-sites
• Performance equal to InfiniBand-level back-end fabric
• Fastest time to first token with optimized workload performance – achieved with end-to-end optimization and minimal fine-tuning
• A converged AI networking solution on a single, unified fabric – AI front-end, back-end, and storage networking
• High-performance support for any GPU (Nvidia or AMD) or AI accelerator, any optics, and any NIC

Want to learn even more about our technology? Join the panel The Infrastructure Race – Powering GenAI at Scale with our own Inbar Lasser-Raab, December 11 at 11:30 AM. 

Throughout the event, DriveNets AI networking experts will be available to answer questions surrounding AI networking, including:

How does DriveNets help optimize CPMT (Cost per Million Tokens)?
DriveNets reduces CPMT with 2 main benefits:

• TCO reduction. Thanks to an open, diversified supply chain, the total network infrastructure cost is reduced and therefore, less CAPEX is applied to each token.
• Higher utilization. With the highest performance networking solution, the entire AI infrastructure is utilized at the highest possible level. Thus, for a given infrastructure, more tokens are processed per any given work duration compared to any other networking solution, and since the ongoing OPEX remains unchanged, CPMT reduces.

How does DriveNets deliver the fastest time to first token?
The firm’s fabric-scheduled Ethernet solution provides a predictable lossless cluster back-end network and minimum deployment fine-tuning to ensure optimized AI workload performance.

In addition, DriveNets Deployment Infrastructure Services (DIS) supports AI cluster architecture design and end-to-end physical installations that accelerate deployment time while maximizing infrastructure deployment.

What is a unified fabric AI networking solution? 
DriveNets Network Cloud-AI delivers a unified networking solution that supports AI back-end fabric, front-end fabric, and storage networking.

Back-end networking in AI clusters refers to the interconnect infrastructure that facilitates internal communication between AI GPUs within a data center. Network Cloud-AI leverages fabric-scheduled Ethernet, an advanced architecture that ensures lossless and predictable network performance while optimizing load balancing and latency. By eliminating packet loss and minimizing GPU idle time, DriveNets Network Cloud-AI optimizes job completion time (JCT), outperforming any other Ethernet or InfiniBand technologies.

Storage networking in AI clusters is responsible for handling the massive data transfer between the AI compute nodes and storage systems. For AI workloads, unlike typical HPC implementations, this is a critical infrastructure, as this traffic is intense and insufficient performance of the storage fabric will result in poor overall workload performance.

DriveNets Network Cloud-AI provides a unified solution for both networking fabrics, sharing the same Ethernet-based technology, architecture, and actual implementation with the back-end fabric.