Bentaus, an Energy-First NeoCloud, Deploys DriveNets High-Performance AI Fabric with AMD GPUs to Produce AI Tokens with 30% Less Energy

DriveNets, a leading provider of high-performance AI networking solutions, and Bentaus, an energy-first NeoCloud, announced a joint solution that drives more tokens from every data center GPU at greater efficiency per watt, and runs the compute as a flexible load, helping solve the power constraint issues of today’s grid. Bentaus cut GPU power 30% while keeping 97% of token throughput on its AI cluster running AMD Instinct™ MI325X GPUs, DriveNets AI Fabric and Bentaus Ziani software.

Read more – Orca Security Expands Platform to Secure Shadow AI

Bentaus develops, deploys, owns, and operates AI infrastructure, hardware, and software. It has deployed DriveNets AI Fabric as the back-end fabric for a new AI cluster running AMD Instinct™ MI325X GPUs. The new cluster works with Bentaus Ziani Systems platform GPU-FLX™ and CPU-FLX™ that manages compute power in real time for enterprises, data centers, and NeoClouds to deliver a grid-responsive, vendor-agnostic AI infrastructure built on open standards that maximizes tokens per watt and lowers cost per token.

Read more – AttackIQ and FRC Advance Federal Cyber Readiness

DriveNets Ethernet-based AI fabric maximizes the performance of AMD Instinct-based AI clusters by performing full-stack networking optimization — from switch, via NIC, to RCCL and kernel optimization. Purposely optimized for AMD environments like the one deployed at Bentaus and based on reference architecture jointly engineered by DriveNets and AMD, it drives GPUs to higher utilization, maximizing tokens-per-second while cutting power consumption per million tokens. In published benchmarks the architecture delivers approximately 5% higher throughput and 12% better time to first token than publicly available results while minimizing power consumption and lowering effective energy costs.

Project timing addresses a live industry pressure point: more than a dozen U.S. states have introduced data center moratoriums or paused legislation in 2026 amid concerns over grid strain and ratepayer costs. New York has already enacted a statewide halt on new large-load permits. Bentaus designs datacenters and AI solutions to be good stewards of the community. Ziani enables data centers to curtail power in real time and provide power back to the grid by dynamically reducing GPU and CPU power consumption while keeping AI workloads running continuously.

“Everyone is racing to add GPUs. We built Bentaus to add compute while mitigating risks to the grid. Ziani flexes our clusters acknowledging grid conditions in real time, and DriveNets’ fabric makes sure the GPUs we run are never waiting on the network. That is how you scale AI capacity now: more compute, delivered more efficiently, on power the grid can sustain.”

Read more – Bitdefender Launches AI Browser Security Advisor

– said Bob Davidoff, CEO and Founder, Bentaus

“The first AI race was to produce more tokens. The next race is to produce them on less power. With AMD and Bentaus, we have built the answer as a complete, open system: our Ethernet fabric drives every GPU to higher utilization and more tokens per watt, and Bentaus’s Ziani keeps that compute running on power the grid can sustain. This is the open, multi-vendor alternative that NeoClouds have been asking for — top-tier performance-per-watt with no lock-in to a single vendor’s stack.”

– said Ido Susan, DriveNets CEO and Founder

“Two shifts are reshaping AI infrastructure at the same time. Buyers are moving from single-vendor stacks to multi-vendor clusters, and power availability is replacing GPU supply as the main limit on how fast capacity comes online. What stands out in this deployment is that the network and the power layer are tuned together, getting the most tokens out of every GPU and every watt. This is quickly becoming the clearest way to compete.”

– said Sameh Boujelbene, VP Data Center Switch and AI Networks Market Research, Dell’Oro Group