• Dell AI Nvidia Blackwell infrastructure promises 4x faster training.
• New PowerEdge supports up to 256 Nvidia GPUs per rack, air and liquid cooling options.

Dell Technologies unveiled its next generation of AI-focused servers powered by Nvidia’s cutting-edge Blackwell Ultra GPUs, an advance for enterprise AI infrastructure. The May 19 announcement at Dell Technologies World in Las Vegas expands the company’s AI Factory (Nvidia’s term for data centres) partnership with Nvidia, promising up to four times faster AI training.

Dell’s enterprise AI infrastructure lineup includes air-cooled PowerEdge XE9780 and XE9785 servers for traditional data centres, and liquid-cooled XE9780L and XE9785L variants for rack-scale deployment. The new systems can be configured with up to 192 Nvidia Blackwell Ultra GPUs with direct-to-chip liquid cooling, expandable to 256 GPUs per Dell IR7000 rack.

According to Dell, these platforms deliver up to four times faster large language model (LLM) training compared to previous generations. “Our job is to make AI more accessible. With the Dell AI Factory with Nvidia, enterprises can manage the entire AI lifecycle in use cases, from deployment to training, at any scale,” said Michael Dell, chairman and chief executive officer.

The enterprise AI infrastructure announcement comes as organisations move from AI experimentation to implementation, creating demand for accessible AI technologies and expertise. Dell’s positioning as “the world’s top provider of AI-centric infrastructure” is timed as companies deploy AI capabilities and attempt to manage the technical challenges involved.

Critical analysis of Dell’s enterprise AI infrastructure strategy

While Dell’s enterprise AI infrastructure advances are impressive on paper, several factors will determine their actual market impact. The company has not disclosed pricing for the high-end systems, which will likely represent significant capital investments for organisations.

The cooling requirements alone, particularly for the liquid-cooled variants, may necessitate substantial data centre modifications for many potential customers. Industry analysts note that Dell faces stiff competition in the enterprise AI infrastructure space from companies like Super Micro Computer, which has also pursued the AI server market aggressively.

However, Super Micro has recently faced production cost challenges and margin pressure, potentially giving Dell an advantage if the latter can maintain competitive pricing.

Jensen Huang, founder and CEO of Nvidia, highlighted the strategic importance of these systems: “AI factories are the infrastructure of modern industry, generating intelligence to power work in healthcare, finance and manufacturing. With Dell Technologies, we’re offering the broadest line of Blackwell AI systems to serve AI factories in clouds, enterprises and at the edge.”

Beyond servers: A comprehensive enterprise AI infrastructure approach

Dell’s enterprise AI infrastructure strategy extends beyond servers, including networking, storage, and software components essential for AI deployment:

• The company has expanded its networking portfolio with the Dell PowerSwitch SN5600 and SN2201 Ethernet switches (part of Nvidia’s Spectrum-X platform) and Nvidia Quantum-X800 InfiniBand switches, delivering up to 800 gigabits per second throughput.
• Dell’s AI Data Platform received upgrades to improve data management for AI applications, including a denser ObjectScale system with Nvidia BlueField-3 and Spectrum-4 networking integrations.
• Dell now offers the Nvidia AI Enterprise software platform directly, including Nvidia NIM, NeMo microservices, and Blueprints.

The company also introduced Managed Services for its AI Factory with Nvidia, providing 24/7 monitoring, reporting, and maintenance.

Availability and market implications

Dell’s enterprise AI infrastructure rollout will be staggered throughout 2025:

• Air-cooled PowerEdge XE9780 and XE9785 servers will be available in the second half of 2025
• Liquid-cooled variants are expected later this year
• The PowerEdge XE7745 server with Nvidia RTX Pro 6000 Blackwell Server Edition GPUs will arrive in July 2025
• The PowerEdge XE9712, featuring GB300 NVL72, will launch in the second half of 2025

Dell is also planning support for Nvidia’s Vera CPU and Vera Rubin platform.

Market outlook for enterprise AI infrastructure

Dell’s push into enterprise AI infrastructure reflects the company’s strategic pivot to capitalise on the AI boom via its established enterprise relationships. As organisations increasingly discover the complexity and expense of implementing AI at scale, Dell appears to be positioning itself as a one-stop solution provider rather than just a hardware vendor.

However, the success of Dell’s enterprise AI infrastructure initiative will depend largely on how effectively AI systems can demonstrate tangible business value beyond raw performance metrics. Organisations investing in such high-end infrastructure will expect clear pathways to operational improvements and competitive advantages.

The partnership with Nvidia gives Dell access to the latest AI accelerator technology, but also creates dependency on Nvidia’s supply chain and roadmap. Given ongoing chip shortages and high demand for AI accelerators, Dell’s ability to secure adequate GPU supplies will be important to meet customer demand.