Dell and Nvidia Unveil AI Factory 2.0 at Conference

CryptoBriefing reports Dell CEO Michael Dell and Nvidia CEO Jensen Huang will deliver a joint keynote at Dell Technologies World on May 18, 2026, with the presentation slated for 10 a.m. PT. CryptoBriefing says the keynote will center on the Dell AI Factory initiative and the launch of Dell AI Factory 2.0, which the article describes as including Blackwell GPUs and new liquid cooling systems. CryptoBriefing reports the companies will make Nvidia AI Enterprise available directly through Dell and that Nvidia-led breakout sessions on AI factories and data activation are scheduled for May 18 and 19.

ITPro provides product-level detail from prior and recent Dell announcements, reporting an expanded PowerEdge server line that includes the air-cooled XE9780 and XE9785 and liquid-cooled XE9780L and XE9785L. ITPro reports those systems support up to 192 Nvidia Blackwell Ultra GPUs with direct-to-chip cooling, and that configurations can be customized to 256 Nvidia Blackwell Ultra GPUs per Dell IR7000 rack. ITPro also reports Dell claims these servers can deliver up to four-times faster large language model training compared with the PowerEdge XE9680. ITPro further reports the PowerEdge XE9712 will include the Nvidia GB300 NVL72, which Dell says offers 50 times more AI reasoning inferencing output, and that the PowerEdge XE7745 will offer the Nvidia RTX Pro 6000 Blackwell Server Edition when it ships in July 2025. ITPro includes a vendor quote from Varun Chharba, SVP of infrastructure and telecom marketing at Dell: "This is a great addition to the AI factory with Nvidia that Dell has, because it's going to really help customers ... adopt AI servers faster than ever before."

Editorial analysis - technical context: Public reporting frames the announcement around three technical vectors common to modern AI infrastructure rollouts: high GPU density per rack, improved cooling (including liquid cooling), and closer bundling of vendor software stacks with hardware. The cited hardware numbers, 192 GPUs per system and 256 GPUs per IR7000 rack, follow a recent industry trend toward extreme density nodes designed for large-model training. Vendor performance claims such as four-times faster LLM training and 50x inferencing gains are presented by Dell via ITPro; independent benchmarks and third-party validation will be required to verify those numbers in production workloads.

Vendors have increasingly moved from selling discrete components to offering integrated hardware-software-service bundles to reduce customer integration burden. Reporting frames the Dell-Nvidia collaboration as an extension of that pattern: Dell brings server, storage, and services, while Nvidia provides accelerator architectures and software like Nvidia AI Enterprise. CryptoBriefing notes Michael Dell has publicly argued that AI's future follows data, implying a focus on where data resides across cloud, on-prem, and edge environments; that framing aligns with the product strategy described in the reporting.

• •Availability and timing: CryptoBriefing reports some new server hardware is expected to launch in July 2025; confirm actual shipment dates and regional availability.
• •Independent benchmarks: look for third-party training and inference benchmarks that validate the four-times and 50x vendor claims.
• •Software and managed services: monitor when Nvidia AI Enterprise images and managed-service SLAs appear in Dell's catalog and whether pricing is disclosed.
• •Customer deployments: vendor demos and case studies at and after the event will show how the integrated stacks perform against end-to-end enterprise requirements.