At Dell Technologies World, Dell Technologies announced significant innovations in their Dell AI Factory in partnership with NVIDIA, designed to accelerate AI adoption and deliver faster time-to-value for enterprises. These advances address the growing demand for scalable, efficient, and accessible AI infrastructure as businesses transition from experimentation to full-scale AI implementation.
Dell AI Factory Infrastructure Innovations
- Next-Gen Compute Servers:
- Dell PowerEdge XE9780 and XE9785 air-cooled servers simplify integration in enterprise data centers.
- Liquid-cooled XE9780L and XE9785L servers support up to 192 NVIDIA Blackwell Ultra GPUs, scalable to 256 GPUs per Dell IR7000 rack, enabling up to 4x faster large language model training.
- PowerEdge XE9712 with NVIDIA GB300 NVL72 delivers 50x AI reasoning inference output and 5x throughput improvements with enhanced power efficiency via Dell PowerCool technology.
- The upcoming XE7745 server, launching July 2025, supports NVIDIA RTX Pro 6000 Blackwell GPUs for physical and agentic AI applications including robotics and digital twins.
- Upcoming Support for NVIDIA Vera CPU and Rubin Platform:
- Dell plans to integrate the NVIDIA Vera CPU and Vera Rubin platform to further enhance performance and efficiency in AI workloads.
- Networking Enhancements:
- Expansion of Dell’s networking portfolio includes the PowerSwitch SN5600 and SN2201 Ethernet switches and NVIDIA Quantum-X800 InfiniBand switches, providing up to 800 Gbps throughput with low latency, supported by Dell ProSupport and Deployment Services.
Dell AI Data Platform Enhancements
- Optimized Data Management:
- Dell ObjectScale introduces a denser, software-defined storage system supporting large-scale AI deployments while reducing costs and data center footprints.
- Integration with NVIDIA BlueField-3 and Spectrum-4 networking enhances performance and scalability.
- New PowerScale, Project Lightning, and PowerEdge XE server solutions leverage KV cache and NVIDIA NIXL libraries for large-scale distributed inference workloads.
- ObjectScale supports S3 over RDMA, achieving up to 230% higher throughput, 80% lower latency, and 98% less CPU load for superior GPU utilization.
- Accelerated Insights and AI Applications:
- Integrated solutions incorporating the NVIDIA AI Data Platform speed up curated data insights and empower agentic AI applications and tools.
Software and Managed Services for AI Deployment
- NVIDIA AI Enterprise Software Platform:
- Available directly from Dell, this platform enables innovation on the Dell AI Factory with NVIDIA, including microservices, reasoning models, and agentic workflow development.
- Red Hat OpenShift support adds flexibility and security to business-critical AI deployments.
- Dell Managed Services:
- New managed services simplify AI operations by handling the full NVIDIA AI solutions stack, offering 24/7 monitoring, updates, and expert support to overcome resource constraints and accelerate outcomes.
Leadership Perspectives
Michael Dell, Chairman and CEO of Dell Technologies, emphasized the mission to democratize AI access:
“Our job is to make AI more accessible. With the Dell AI Factory with NVIDIA, enterprises can manage the entire AI lifecycle across use cases, from training to deployment, at any scale.”
NVIDIA CEO Jensen Huang highlighted the strategic importance of AI factories:
“AI factories are the infrastructure of modern industry, generating intelligence to power work across 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.”
Dell Technologies and NVIDIA’s collaborative innovations in compute, storage, networking, software, and managed services underscore their commitment to empowering enterprises with scalable, efficient AI infrastructure. The Dell AI Factory with NVIDIA delivers end-to-end solutions that streamline AI adoption from experimentation to production, helping businesses accelerate AI-driven transformation across industries.
