At Dell Technologies World 2025, Dell unveiled significant enhancements to its AI Factory portfolio, reinforcing its position as a leading provider of enterprise AI infrastructure. The updates span edge to data center environments and emphasize scalable deployments, improved energy efficiency, and broad partner integrations. Key announcements include high-performance PowerEdge servers equipped with AMD Instinct MI350 GPUs, new energy-saving cooling technologies, and Dell AI Factory integrations with Intel Gaudi 3, NVIDIA, and a range of AI software partners.
At the heart of Dell’s data center advancements are the PowerEdge XE9785 and XE9785L servers, now supporting AMD’s MI350 GPUs with 288GB of HBM3E memory per accelerator. Designed for both air- and liquid-cooled environments, these systems offer up to 35x higher inferencing performance and reduced cooling energy costs. To further optimize dense AI workloads, Dell introduced the PowerCool Enclosed Rear Door Heat Exchanger (eRDHx), a self-contained system that captures 100% of IT-generated heat and reduces cooling energy use by up to 60%. Additionally, Dell’s Linear Pluggable Optics and upgraded AI Data Platform—including Project Lightning and an enhanced Lakehouse—aim to reduce latency, cut power consumption, and accelerate AI pipeline workflows.
Dell also expanded its AI Factory ecosystem with support for Intel Gaudi 3 AI accelerators and partnerships with Meta (Llama 4), Mistral AI, Google (Gemini), Cohere, and Glean. These integrations enable enterprises to deploy full-stack AI platforms, run large language models, and manage AI agents on-premises. For edge use cases, the Dell Pro Max Plus AI PC incorporates an enterprise-grade NPU capable of inferencing 109B-parameter models locally. Across its portfolio, Dell emphasizes open, modular, and secure solutions with support for both on-premises and hybrid deployments, aiming to lower the cost of inferencing by up to 62% compared to public cloud alternatives.
• Dell PowerEdge XE9785 and XE9785L support AMD MI350 GPUs with 288GB HBM3E each
• New PowerCool eRDHx system reduces data center cooling costs by up to 60%
• Dell Project Lightning delivers 2x throughput over other parallel file systems
• Intel Gaudi 3 and AMD ROCm stack supported in new Dell AI Platform integrations
• New Dell AI PC with Qualcomm AI 100 Card enables edge inferencing of 100B+ parameter models
• Partner integrations include Meta (Llama 4), Mistral AI, Glean, Cohere, and Google Gemini
• Dell AI Factory supports agentic AI, scalable search, and LLM inferencing on-prem
• Expanded services include Dell AI Security and Resilience for full-stack protection
“Our latest AI advancements — from groundbreaking AI PCs to cutting-edge data center solutions — are designed to help organizations of every size to seamlessly adopt AI, drive faster insights, improve efficiency and accelerate their results,” said Jeff Clarke, Chief Operating Officer, Dell Technologies.
