Dell Technologies recently showcased its latest innovations in private cloud, storage, and cyber resilience at a major product demonstration event. During a media briefing, Liu Zhihong, storage business general manager for Dell Technologies' infrastructure solutions division in Greater China, highlighted a growing trend: while intelligent applications are on the verge of explosive adoption, few have achieved large-scale implementation within enterprises.
Liu pointed to a fundamental challenge: how exactly can intelligence be achieved? Whether it's robotic intelligence or agentic AI, reaching or exceeding human-level capability requires massive training cycles. While algorithms and computing power are readily available from market suppliers, the true ceiling on intelligence is determined by data. He drew an analogy comparing new university graduates entering a company, noting they often can't immediately perform at full capacity because every enterprise has its own technical know-how. "Even with identical computing power and identical models, the level of intelligence ultimately developed will differ," he said. This underscores a critical reality: compared to algorithms and compute, the quality and volume of a company's own proprietary data is what truly determines its AI capability. This irreplaceable asset is the real bottleneck for intelligent applications, setting the limit on what level of intelligence can actually be achieved.
However, enterprise data is often scattered across systems built over the past few decades. ERP, CRM, and email platforms operate in silos, while vast amounts of critical information remain buried in engineers' laptops, individual emails, or meeting minutes. Liu noted that "this scattered data often contains the most important know-how of a company." Additionally, larger datasets become harder to migrate, and data existence doesn't guarantee usability — data quality remains equally essential.
To address these challenges, Dell Technologies introduced its Intelligent Data Platform, which transforms fragmented raw enterprise data into high-performance data foundations that can be directly invoked by intelligent applications. By leveraging storage engines, data engines, and data orchestration engines, the platform establishes a data pipeline between datasets and applications. Xu Liangmou, enterprise technology strategy director for Dell Technologies' infrastructure solutions division in Greater China, outlined two distinguishing features of the solution. First is real-time awareness of data changes. In traditional setups, enterprises rely on manual checks or batch processing to identify new or modified files, causing delays where front-end systems can't access the latest information in a timely manner. "Now we can sense data changes at the backend and help front-end intelligent applications retrieve the latest data status in real time."
The second feature is enhancing long-context support through storage. Intelligent applications often demand long context windows, but GPU memory is limited. Technologies like KV Cache offloading transfer part of the context caching pressure to the storage system, easing the memory bottleneck. The surge in intelligent applications has also triggered explosive demand for storage and other infrastructure. According to IDC's first-quarter 2026 Global Enterprise Storage Systems Tracker, Dell Technologies Inc. (DELL) leads the market with a 31.2% share. This dominant position stems from over four decades of storage product expertise accumulated by the company.
Liu also highlighted the latest upgrade to Dell's PowerStore Elite all-flash storage line. Its core data reduction guarantee has improved from 5:1 last year to 6:1 this year, delivering up to three times higher IOPS, storage density, and throughput. The solution also supports non-disruptive upgrades, providing stable, efficient data handling capacity for enterprises facing continuously growing data demands and intelligent workloads.