Dell Unveils Nvidia Blackwell-Powered Servers for AI
Dell Technologies has announced updates to its Dell AI Factory with Nvidia by introducing new infrastructure, software and managed services aimed at organisations moving to enterprise-wide implementation.
Its new PowerEdge XE9785L servers can support up to 256 Nvidia Blackwell Ultra GPUs per rack, as technology firms collaborate on infrastructure solutions. The company is introducing these products alongside improvements to its data platform and new management services, at a time when businesses face mounting pressure to operationalise AI technologies.
“We're on a mission to bring AI to millions of customers around the world,” says Michael Dell, Chairman and CEO of Dell Technologies.
“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.”
Delivering significant performance increases with Nvidia Blackwell
The newly announced compute solutions include both air-cooled PowerEdge XE9780 and XE9785 servers, which are designed to integrate with existing enterprise data centre infrastructure and liquid-cooled PowerEdge XE9780L and XE9785L servers for rack-scale deployments.
Additionally, the liquid-cooled variants support up to 192 Nvidia Blackwell Ultra GPUs with direct-to-chip cooling technology, while rack configurations can accommodate up to 256 Nvidia Blackwell Ultra GPUs per Dell IR7000 rack.
Dell reports these platforms are able to deliver four times faster large language model (LLM) training capabilities when utilising the 8-way Nvidia HGX B300.
- New PowerEdge servers deliver up to 4x faster LLM training with 8-way Nvidia HGX B300
- Dell ObjectScale with S3 over RDMA achieves 230% higher throughput and 80% lower latency
- Dell Managed Services offers 24/7 monitoring and management of the full Nvidia AI stack
For reasoning tasks, Dell has introduced the PowerEdge XE9712 with Nvidia GB300 NVL72, which the company states provides 50 times more AI reasoning inference output and five times improvement in throughput.
This platform incorporates Dell PowerCool technology to help businesses improve their power efficiency moving forward – which is important, as data centres are being urged to be more sustainable.
The company has also confirmed plans to support Nvidia's Vera CPU and Vera Rubin platform with a forthcoming PowerEdge XE server designed specifically for Dell Integrated Rack Scalable Systems.
Improving high-performance AI data access
The updated Dell AI Data Platform focuses on providing continuous data access for AI applications with enhancements to the ObjectScale storage system. Dell says that these improvements support large-scale AI deployments while reducing data centre footprint through a denser, software-defined architecture.
Now, the platform integrates with Nvidia BlueField-3 and Spectrum-4 networking to address performance and scalability requirements.
Dell has revealed ObjectScale will support S3 over RDMA protocol, achieving 230% higher throughput, 80% lower latency and 98% reduced CPU load compared to traditional S3 implementations, resulting in improved GPU use.
Likewise, Dell has created an integrated solution combining PowerScale storage, Dell Project Lightning and PowerEdge XE servers for high-performance computing (HPC) tasks.
By integrating with the Nvidia AI Data Platform, Dell hopes to accelerate data processing for agentic AI applications.
To support the networking requirements, Dell expanded its portfolio to include the PowerSwitch SN5600 and SN2201 Ethernet switches, which are components of the Nvidia Spectrum-X Ethernet networking platform, alongside Nvidia Quantum-X800 InfiniBand switches.
These high-density switches provide 800 gigabits per second throughput and receive support through Dell ProSupport and Deployment Services.
Dell Managed Services for Nvidia AI Factory: Addressing skills gaps
To address the technical expertise challenges in enterprise AI deployment, Dell launched Managed Services for the Dell AI Factory with Nvidia.
This offering aims to provide management for the complete Nvidia AI solutions stack, including AI platforms, infrastructure and Nvidia AI Enterprise software. It seeks to deliver 24/7 monitoring, reporting, version upgrades and patching, targeting organisations experiencing resource and expertise constraints in their AI implementation efforts.
Software enhancements include direct availability of the Nvidia AI Enterprise software platform from Dell, giving organisations access to Nvidia NIM, Nvidia NeMo microservices, Nvidia Blueprints, Nvidia NeMo Retriever for RAG and Nvidia Llama Nemotron reasoning models on the Dell AI Factory with Nvidia.
Dell also added Red Hat OpenShift availability on the Dell AI Factory with Nvidia to simplify business-critical AI deployments while maintaining flexibility and security requirements.
These announcements from Dell are reflective of changing market dynamics across the data centre industry, particularly as organisations progress from AI experimentation to implementation.
“AI factories are the infrastructure of modern industry, generating intelligence to power work across healthcare, finance and manufacturing," says Jensen Huang, Founder and Chief Executive Officer of Nvidia.
“With Dell Technologies, we're offering the broadest line of Blackwell AI systems to serve AI factories in clouds, enterprises and at the edge.”
Explore the latest edition of Data Centre Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
Subscribe to our Weekly Newsletter for the latest insights, trends and expert analysis.
Data Centre Magazine is a BizClik brand
