Robin Braun, vice president of AI business development at Hewlett Packard Enterprise, put it bluntly. “People are trying to find the signal in the noise.” That search for signal is driving a rethinking of how AI infrastructure should be built. Braun spoke with theCUBE for HPE’s “Unleash AI Momentum” series.
The message: the inference era is here. And it demands a different kind of data center.
HPE and Kamiwaza Corp. are betting on a mix. Not just graphics processing units. Not just central processing units.
Both. The idea is to deliver a full set of AI computing solutions, from hosting applications to generating intelligence, from static workflows to agentic orchestration systems.
Think of it as a factory floor for data. Braun described organizations struggling to get past the noise surrounding the IT stack. The goal is to use infrastructure to improve inference speed.
That speed is the bottleneck for making sense of vast amounts of information. “That’s where inference comes in,” Braun said. “Just trying to use that to get at the data.” The practical upshot is production-ready enterprise AI.
Companies want governance. They want security.
They want scale. They want to move from pilot to production without falling flat. Many organizations are failing to unlock the full potential of their data.
That is the core challenge. Braun emphasized using data to improve efficiency and business outcomes.
Inference is the tool for that job. The shift is structural. The data center of the future looks less like a server room and more like an AI factory.
Static workflows are giving way to agentic orchestration systems. These systems don’t just process data. They act on it.
HPE and Kamiwaza are positioning themselves for that future. The inference era requires infrastructure that can handle both the heavy lifting of training and the rapid, repeated queries of inference.
A single chip architecture won’t cut it. Braun’s comments point to a fundamental tension. Organizations are drowning in data.
They are buried in noise. The IT stack itself generates much of that noise.
Companies need a way to cut through it. Inference is the filter. It turns raw data into actionable insight.
But inference requires speed. If the infrastructure lags, the signal gets lost. The partnership between HPE and Kamiwaza is a bet that the market is ready for this shift.
Enterprises are no longer experimenting with AI. They are deploying it.
And deployment demands a different set of tools. Governance and security are not afterthoughts. They are prerequisites.
Braun’s comments suggest that companies are learning this the hard way. Pilots work.
Scaling them fails. The reason is often infrastructure. The hardware is wrong.
The architecture is wrong. The balance between CPU and GPU is off. HPE and Kamiwaza are offering a rebalance.
A mix of platforms. A full set of solutions.
The goal is to make inference fast enough to be useful. Braun did not promise easy answers. He described a struggle.
Organizations are trying to use their data to improve efficiency. They are trying to improve their business.
Inference is the mechanism. But the mechanism only works if the infrastructure supports it. That is the bet HPE and Kamiwaza are making.
The inference era is not a future state. It is the present. And the present demands a rethink.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “HPE and Kamiwaza target inference era with mixed AI data center approach”, “description”: “Robin Braun, vice president of AI business development at Hewlett Packard Enterprise, put it bluntly. “People are trying to find the signal in the noise.”…”, “datePublished”: “2026-06-22T07:51:00+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Sebastian Smith”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.2bme.xyz/”}}
