Dell unveils AI PCs with Nvidia GPUs for AI model testing

New PCs introduced Tuesday by Dell sport high-powered Nvidia GPUs designed to help workers test AI models before deployment.

The new AI PC lineup includes laptops and desktops with Nvidia’s latest Blackwell Ultra and Blackwell GPUs, which provide the muscle needed to test compute-intensive AI models. AI PCs are designed to run generative AI tools and models more efficiently than standard PCs.

“These AI developer PCs will make it much easier for developers to prototype, to test and even scale their models into production environments,” said Kevin Terwilliger, vice president and general manager for consumer, commercial and gaming PCs at Dell.

The announcement coincides with Nvidia’s rollout of its latest GPU Blackwell Ultra at the GTC conference in San Jose. The GPU is the successor to the red-hot Blackwell GPU, which racked up $11 billion in sales in the most recent quarter.

The Pro Max series of workstations bring hardware previously used in data centers to desktops. “These systems will redefine AI developer experiences by really bridging the gap between desk side experimentation and enterprise scale AI deployment,” Terwilliger said.

The new Pro Max with GB300 packs the tightly coupled Blackwell Ultra GPU and Grace CPU, both made by Nvidia. It includes 784GB of unified system memory — 288GB HBM3e memory and 496GB of LPDDR5X memory.

Developers can test and prototype AI models with up to 460 billion parameters on the Pro Max with GB300. The system delivers 20 petaflops of performance with the FP4 data type, which is a low-precision measure for inferencing.

The Pro Max with GB10 includes Nvidia’s Blackwell GPU — a generation behind the Blackwell Ultra — and the Grace CPU. Developers will be able to work with AI models with up to 200 billion parameters on a  PC that delivers 1 petaflop of FP4 performance.

The systems are slated to be available later this year, include Nvidia’s DGX Base OS or Ubuntu Linux, and will be preconfigured with Nvidia’s AI software.

Desktop prototyping makes it easy to determine the size and performance of AI models, according to Terwilliger. “Once you right size it for the data set, then it’s appropriate to then scale it into the data center,” he said.

Some Dell customers need a secure test environment where they can tune some of their own data, said Matt Toolan, a Dell spokesman. For example, customers have tested different digital assistants to determine whether chat capabilities and responses were meaningful and on target, Toolan said.

Anshel Sag, principal analyst at Moor Insights and Strategy, was skeptical about Dell’s claim of using Pro Max PCs to scale AI models to data centers.

Instead, the AI tools more about making the deployments of on-device AI smoother, Sag said. “I don’t really think PC workloads are really useful for data-center scale deployments. I do believe Dell is uniquely positioned to enable hybrid AI that are entirely on-premise,” Sag said.

Dell also announced Pro Max laptops with RTX Pro GPUs based on the Blackwell architecture. The company claimed the PCs will be up to 36% faster than the previous-generation Precision PCs.

The Dell Pro Max laptops have “tandem” OLED displays, haptic touchpads, lattice-less keyboards and 8-megapixel IR cameras. The systems come with Intel Core Ultra Series 2 or AMD Ryzen processors, offer screen sizes from 14-in. to 18-in., and will ship later this year. 

Dell did not provide pricing information for the systems.