Ziani Systems retweeted
Bentaus deploys DriveNets High-Performance AI Fabric with AMD GPUs to produce AI Tokens with 30% Less Energy
drivenets.com/news/bentaus-d…
Paired with Bentaus Ziani orchestration software, the deployment offers flexible GPU capacity that balances the grid instead of working against it, cutting energy per token by 30% at a cost of only 3% token throughput
Join @drivenets and @BentausOfficial at SC26 for a presentation of our joint solution in the @AMD booth (3028) on Thursday, November 19th at 11:30 AM
Ziani Systems retweeted
30% less energy per token.
How did this energy-first NeoCloud maximize AI performance per watt.
@BentausOfficial built its 2,560-GPU AMD Instinct™ MI325X cluster with DriveNets AI Fabric and its Ziani power orchestration software, achieving 30% less energy per token with minimal impact on cluster performance.
For an energy-first NeoCloud, every watt has to turn into tokens.
@drivenets AI Fabric gave Bentaus the network side: a full-stack-optimized Ethernet fabric on a validated @AMD Instinct reference architecture that keeps GPUs productive and the stack open. On that foundation, Ziani dials cluster power down and up with the grid while workloads keep running, delivering AI capacity that is high-performing, efficient per watt and sustainable on the grid.
👉 Read the case study drivenets.com/resources/neoc…
Ziani Systems retweeted
Power has become the main limit on how fast AI capacity can come online.
Our answer, built with @DriveNets and @AMD, tunes the network and the power layer together. DriveNets AI Fabric drives GPUs to higher utilization, and Bentaus’ @ZianiSystems flexes compute power in real time so AI infrastructure can respond to the grid.
Ziani Systems retweeted
Thank you, @arm!
Our @ZianiSystems integration with Arm Neoverse happened because of the people at Arm who worked with us on it. We’re now working on the Arm AGI CPU, and we’re excited to share those results as well. Stay tuned!
The CPU revolution is here, it runs on Ziani by @BentausOfficial. At #OCPGlobalSummit, we're showing CPU-FLX™ with @Arm- dynamic CPU frequency and power control that turns compute into a grid-responsive load. @bobdavidoff and Kevin Matcham will be at the booth.
📍 Arm booth, San Jose
📆 Tue & Wed, 2–4 PM PT
Ziani Systems retweeted
More detail on the Bentaus side of the @Synota_ partnership:
@ZianiSystems was built to read power in real time. Modulating GPU load against grid dispatch signals only works if the orchestration layer knows actual draw as it happens, which is why the control loop includes a telemetry and validation stage that confirms execution before anything gets reported back to the grid.
That same telemetry is what the settlement framework will draw on. Ziani supplies the measurement of what the infrastructure actually consumed. Synota supplies the billing and settlement technology that connects that measurement to customer commercial terms.
The framework is being designed to cover GPU, CPU, token, energy and AI storage usage, with daily and monthly settlement first.
First deployment is targeted for Q4 2026 on Bentaus infrastructure.
Ziani will play an integral role in @Bentaus’ collaboration with @Synota_.
Our DCIM telemetry and Power Asset Orchestrator will provide transparency into GPU, token and energy usage to support settlement, while enabling AI factories to operate as flexible loads.
Another step in building rich software for AI infrastructure and NeoClouds.
Bentaus has strategically selected @Synota_ as the settlement partner for the Bentaus #NeoCloud.
Together, we’re developing settlement for AI infrastructure across GPU, CPU, token, energy and storage usage, powered by telemetry from @ZianiSystems.
Daily + monthly settlement, moving toward next-generation token-based settlement for AI compute.
Full release: world.einnews.com/pr_news/93…
26
🖖 100% this
Luxor Energy and @BentausOfficial recently completed the first ever AI GPU response to ERCOT 4CP signals:
16
Ziani Systems retweeted
How do you get a GPU running AI inference to respond to the grid?
One integrated loop: ERCOT conditions → Luxor's QSE signal → @ZianiSystems orchestration → GPU curtails to ~25%. Under half a second, no lost work.
The full picture 👇
Ziani Systems retweeted
How do you get a GPU running AI inference to respond to the grid? Check out how we worked with @luxor using our @ZianiSystems
How do you get a GPU running AI inference to respond to the grid?
One integrated loop: ERCOT conditions → Luxor's QSE signal → @ZianiSystems orchestration → GPU curtails to ~25%. Under half a second, no lost work.
The full picture 👇
Ziani Systems retweeted
First response of an AI inference GPU to a live @ERCOT_ISO 4CP signal.
@luxor Energy's QSE generated the curtailment signal. @ZianiSystems by @BentausOfficial took the GPU from full draw to roughly 25% in under 500 milliseconds, closed loop and verified, with inference workloads checkpointed, paused, and resumed. Nothing dropped.
Four fifteen minute peaks set a large share of a Texas data center's annual transmission bill. Now AI load can respond to them. ERS enrollment next.
prnewswire.com/news-releases…
Ziani Systems retweeted
Luxor Energy, the energy division of @luxor and @BentausOfficial report that their Ziani platform cut a live GPU's power draw to 25% of normal in under 500 milliseconds, while inference workloads kept running. It's the first demonstrated AI GPU response to an ERCOT 4CP signal.
Why it matters: ERCOT's Four Coincident Peak program rewards large loads that pull back during the grid's highest demand intervals each summer (June through September). Until now, AI data centers have largely sat outside that conversation as fixed, inflexible load. This result suggests that doesn't have to be the case. GPU fleets could become active participants in market programs, not just consumers of power, with real potential to lower transmission costs.
Congrats to the Luxor Energy and Bentaus teams on a genuinely useful proof point for grid-aware AI infrastructure.
luxor.tech/news/corporate-ne…
How is Ziani licensed?
Ziani offers the software as a plug-in for SuperCloud Composer via the PowerAsset™ Orchestration License (SFT-PAO-SINGLE). Ziani also provides an IoT controller license for the Bentaus Edge™ (SFT-BENT-IOT) and a free 30-day trial for up to 200 nodes (SFT-PAO-TRIAL).
26
How does Ziani help with SLA (Service Level Agreement) compliance?
Ziani's frequency-adjusted GPU orchestration ensures that even when power limits are reduced to meet grid signals, the GPUs remain powered, available, and actively processing workloads. Workloads can seamlessly shift across clusters, ensuring SLA uptime commitments are honored even during grid curtailment.
20
How is Ziani's software verified for security and stability?
Ziani Systems is undergoing formal verification of its energy orchestration code through a partnership with Logical Intelligence. This rigorous mathematical auditing process, typically reserved for mission-critical systems like aerospace and blockchain, ensures the code is deterministically proven to execute without hallucination or failure.
1
20
What environments can Ziani's Power Asset Orchestrator (PAO) be deployed in?
Built for enterprise scalability and low latency, PAO supports on-premises, hybrid, and cloud environments.
10
Can the PAO AI Agent act autonomously?
Yes, the AI Agent uses AI-driven business rules to dynamically and autonomously adjust compute loads, shift training jobs, or ramp resources to capture energy savings at the edge, rack, or data center level without requiring human intervention.
13
What does the built-in PAO AI Agent do?
The Power Asset Orchestrator (PAO) AI Agent is an interactive, conversational interface that executes task-oriented operations across both compute and energy infrastructure. Users can prompt the agent to fetch metrics (like PUE or GPU load efficiency) or issue direct commands like "optimize BESS charge cycles" or "reduce cooling by 5%."
14
Does Ziani require new hardware to modulate GPU power?
No. Ziani's GPU-FLX™ utilizes existing DCIM (Data Center Infrastructure Management) standards. It requires no hardware modifications to unlock fast, predictable load flexibility through software orchestration.
13
How does Ziani improve a data center's PUE (Power Usage Effectiveness)?
Power Asset Orchestrator (PAO) provides continuous, real-time PUE insights that trigger intelligent orchestration. If PUE spikes, the software automatically adjusts cooling systems, initiates load shedding, or shifts AI workloads to optimize facility efficiency and lower operating costs.
8



















