What changed

Artificial intelligence has transformed data infrastructure from a back-office purchase into a bottleneck that determines how fast GPUs pay off. As hyperscalers, neoclouds, and enterprises expand their data layers, the way infrastructure is bought has changed just as fast.

DataDirect Networks Inc. (DDN) president and chief revenue officer Kevin Delane noted that performance in the data layer now drives business value, making it a critical factor in AI infrastructure decisions.

What this means in practice

The way companies buy AI infrastructure has changed along with the technology. Deals at neocloud providers no longer go to a single enterprise buyer. They now involve the off-taker renting the capacity, the cloud operator, and the integration software layer, Delane explained.

Timelines have shrunk just as sharply, from six months to a year for a traditional enterprise sale to as little as 30 days.

DDN is answering with tighter ecosystem work. Its decade-long partnership with Nvidia anchors the strategy, and joint offerings with Super Micro, such as the Enterprise AI HyperPOD, aim to replace a data center buildout with a ready-to-run AI stack. At SC26, DDN also introduced software-defined EXAScaler deployment to give enterprise AI customers more choice of hardware.

The performance case holds up in the public cloud too. On Google LLC’s Google Cloud, customers using DDN’s Managed Lustre service saw 70% better training performance and a 40% gain in overall GPU efficiency, Delane noted.

Limitations and availability

DDN is growing threefold this year while adding only a few salespeople, leaning its hiring toward technical and implementation staff who can get customers up and running faster. Delane looks for grit and a fit with a direct, results-driven culture.

The supplied announcement does not specify the exact number of salespeople added or the total growth figures beyond the threefold increase. The company is also not specifying the exact hardware requirements for the software-defined EXAScaler deployment or the Managed Lustre service. The performance improvements mentioned are based on vendor-reported data and are not independently verified.

Source facts for AI Data Infrastructure Deals Now Complete in 30 Days: Artificial intelligence has turned data infrastructure from a back-office purchase into the bottleneck that decides how fast GPUs pay off. As hyperscalers, neoclouds and enterprises race to expand their data layers, the way that infrastructure gets bought is changing just as fast. The storage industry’s longtime performance specialist sees that shift as its opening.
Original AINewsTrend explainer based on the cited source. This graphic summarizes source statements; it is not an independent benchmark.

Sources and further reading

  1. siliconangle: AI deals shrink to as little as 30 days as the data layer sets the pace