The USA Leaders
20 May 2025
Austin – Dell AI Servers are making headlines—and for good reason. In a bold move that signals its intent to reclaim leadership in enterprise computing, Dell Technologies has unveiled its most powerful AI infrastructure yet, featuring Nvidia’s cutting-edge Blackwell Ultra GPUs. At the heart of this announcement is a very clear message to Silicon Valley: Dell is not just back in the AI game—it’s swinging for the fences.
At the 2025 Dell Technologies World event, the company introduced a lineup of next-gen PowerEdge AI servers that promise to redefine the speed, scale, and efficiency with which organizations can train and deploy large language models, generative AI systems, and real-time applications. Here’s a detailed breakdown of how Dell is rebuilding its future around AI—and why it matters to enterprises looking to accelerate digital transformation.
PowerEdge Reimagined: The Flagship Dell AI Servers
Dell’s newest PowerEdge XE9785 and XE9785L servers are the crown jewels of its AI expansion. These high-performance systems come in both air- and liquid-cooled variants and are built for one purpose: to handle massive AI workloads with zero compromise.
These machines are packed with AMD Instinct MI350 GPUs, offering 288GB of HBM3E memory per chip. That translates into up to 35 times more inferencing power than previous-generation servers—an extraordinary leap that positions Dell as a top contender in AI infrastructure.
| Server Model | GPU Support | Cooling Options | Key Features | Availability |
| PowerEdge XE9785/XE9785L | AMD Instinct MI350 Series (288GB HBM3E per GPU) | Air & Liquid-cooled | Up to 35x inferencing performance, GenAI/HPC optimized | 2025 |
| PowerEdge XE9780/XE9780L | Nvidia HGX B300 (Blackwell Ultra) | Air & Liquid-cooled | Up to 192 (custom 256) GPUs, 4x faster LLM training | 2H 2025 |
| PowerEdge XE9712 | Nvidia GB300 NVL72 | Enhanced cooling | 50x AI inference output, 5x throughput, PowerCool tech | 2025 |
| PowerEdge XE7745 | Nvidia RTX Pro 6000 Blackwell | Air-cooled | Universal platform for robotics, digital twins, multimodal AI | July 2025 |
List of Latest Dell AI Servers
Blackwell Ultra Inside: Why Nvidia’s New Chips Are a Game-Changer
The collaboration with Nvidia is where things really get interesting. Dell’s AI servers now support the groundbreaking Nvidia Blackwell Ultra GPUs—hardware engineered to deliver up to four times faster AI model training speeds compared to older systems.
Each Blackwell Ultra GPU packs advanced tensor cores, second-gen transformer engines, and 288GB of lightning-fast memory, making it a dream setup for companies running large-scale AI operations. With support for 192 to 256 GPUs per rack and up to 1.1 exaflops of FP4 performance, Dell’s AI infrastructure is now capable of handling even trillion-parameter models at speed and scale.
Project Lightning: Turbocharging the AI Data Layer
No AI stack is complete without a high-speed data layer, and Dell’s response is Project Lightning—a parallel file system that’s poised to become the fastest in the world. Integrated into the Dell AI Data Platform, it enables everything from vector search and text summarization to SQL-accessible LLM functions—all within a unified Data Lake House environment.
This means faster, smarter, and more seamless AI development for enterprise teams.
Cooling the Future: Sustainable AI at Scale
AI training consumes vast amounts of energy. Dell addresses this challenge head-on with modular, rack-scale cooling solutions. From advanced air cooling to direct-to-chip liquid systems, the servers are built to manage up to 264kW per rack without breaking a sweat.
With disaggregated power shelves, hybrid DLC systems, and up to 480kW shared power support, Dell’s IR7000 racks aren’t just powerful—they’re sustainable and service-friendly.
Built for Tomorrow: Compatibility with Next-Gen Nvidia CPUs
Future-proofing is a major concern for enterprise IT leaders. Dell’s new AI servers are designed to be compatible with Nvidia’s forthcoming Vera and Vera Rubin CPUs, ensuring long-term value and adaptability as AI hardware evolves. The integration with Nvidia Spectrum-X and InfiniBand Quantum-X800 networking delivers up to 800 Gbps throughput, essential for low-latency, high-bandwidth applications.
Blackwell Ultra + Dell: The Impact on Real-Time AI
Dell’s PowerEdge XE9712 servers, loaded with Blackwell Ultra chips, offer up to 50x more AI reasoning output than earlier models. This makes them ideal for real-time AI scenarios—like automated decision-making, digital assistants, or physical robotics—where speed and precision are non-negotiable.
Blackwell’s FP4 format allows for rapid content generation and token processing, up to 10x faster than the prior H100 generation, enabling enterprise-grade performance even for latency-sensitive tasks.
Custom Configurations for Every AI Use Case
Flexibility is baked into Dell’s design. Whether you’re running large-scale distributed training, deploying multi-server clusters, or fine-tuning customer-facing AI services, Dell offers extensive configuration options. From modular OCP-inspired racks to AI Solutions Explorer tools that help tailor hardware to specific workloads, Dell is making AI infrastructure both customizable and accessible.
Strategic Shift: Dell’s AI Factory and Managed Services
Dell isn’t just selling servers—it’s offering a complete AI ecosystem. The AI Factory concept introduced at Dell Technologies World 2025 brings full-stack enterprise AI to customers through managed services, infrastructure bundles, and software tools like NVIDIA NeMo, NIM, and Blueprints. Dell claims this approach can deliver up to 60% cost savings compared to public cloud AI offerings.
In a competitive landscape dominated by the likes of Super Micro and other hyperscalers, this full-stack approach could be Dell’s differentiator.
The Bottom Line: Is Dell Back in the Race?
Absolutely. With the Dell AI Servers lineup, powered by Nvidia Blackwell Ultra GPUs, the tech giant is delivering unmatched performance, energy efficiency, and scale for enterprise AI. While profitability will be a challenge due to rising production costs and competitive pricing pressure, Dell is clearly positioning itself not just as a server vendor but as a key enabler of AI transformation for the modern enterprise.
For businesses navigating the AI arms race, Dell’s latest offerings present a compelling case: powerful infrastructure, forward-compatible technology, and a complete ecosystem ready for deployment.
Also Read: Nvidia’s Taiwan Investment is HUGE With an AI Supercomputer: A Strategic Move for the US?
