Dell Applied sciences has introduced its newest suite of AI infrastructure options constructed upon NVIDIA’s Vera Rubin platform. This growth goals to introduce PowerEdge servers outfitted with the Vera Rubin NVL72 structure, delivering 3.6 exaflops of processing energy.
The PowerSwitch networking options will combine Spectrum-6 expertise, providing 102.4 terabits per second switching capability. This enhancement, coupled with BlueField-4 integration, goals for 800 gigabits per second efficiency, supporting a spread of NVIDIA Nemotron 3 AI fashions.
Dell enhances its AI server line by embedding NVIDIA’s Rubin platform, a transfer highlighted throughout Dell Applied sciences World in 2025. This collaboration goals to advance each agentic AI and AI reasoning capabilities.
Upcoming PowerEdge servers will characteristic the NVIDIA Vera Rubin NVL72 platform, delivering AI efficiency along with 75 terabytes of quick reminiscence. They may also incorporate resiliency capabilities inside Dell’s rack-scale infrastructure.
Increasing HGX Server Capabilities
The enlargement of Dell’s AI server portfolio continues with assist for NVIDIA HGX Rubin NVL8 configurations. These techniques goal for roughly 400 petaflops of AI efficiency, that includes 2.3 terabytes of HBM4 reminiscence and 176 terabytes per second reminiscence bandwidth.
That is designed to supply deployment flexibility for numerous knowledge centre necessities, supporting high-performance AI infrastructure with out requiring main facility adjustments.
Networking & Information Processing Enhancements
Dell’s PowerSwitch portfolio sees progress with the inclusion of NVIDIA Spectrum-6 Ethernet expertise, extending assist to 512 ports of 800G connectivity with co-packaged optics. This targets a 5x energy effectivity enhancements and enhances reliability by 10-fold.
Complementing these networking developments are improvements throughout the Dell AI Information Platform with NVIDIA BlueField-4 knowledge and storage processor integration, providing capabilities in KV Cache storage tiering and AI-native knowledge processing.
Supporting Clever AI Fashions
The combination of the NVIDIA Nemotron household, spanning from 30B to 500B parameters, marks an development in AI intelligence and effectivity. These fashions, with their hybrid mixture-of-experts structure, search to empower organisations to deploy refined reasoning workloads.
As AI infrastructure calls for surge, pushed by cloud and enterprise environments, Dell’s choices, in collaboration with NVIDIA, goal to deal with these challenges by delivering scalable structure options that evolve with organisational wants.
These platforms goal to assist enterprise AI infrastructure, offering expertise designed to assist companies deploy and scale AI functions successfully.
