Nvidia continues its aggressive push into AI infrastructure with multiple announcements spanning open-source models, enterprise software acquisition, and new hardware platforms designed for increasingly demanding AI workloads.
Nvidia launched Nemotron 3 Ultra, a 550-billion parameter mixture-of-experts open model that ranks as the smartest openly available US model, though it trails Chinese competitor Kimi K2.6 according to Artificial Analysis benchmarks. The model uses approximately 55 billion active parameters at any given time.
In parallel, Nvidia acquired Kumo AI for over $400 million, strengthening its enterprise AI software capabilities. Kumo, founded five years ago, provides predictive AI solutions to enterprise customers and raised $37 million at a $250 million valuation in 2022.
On the hardware front, Nvidia unveiled the DGX Station, a desktop Windows PC leveraging its Grace Blackwell GB300 chip with up to 748 GB of memory. The system can run trillion-parameter models, bringing datacenter-class AI capabilities to desktop environments.
The company also announced that its Vera Rubin computing platform is entering full production, with first systems shipping this fall following a March announcement.
Looking ahead, Computex 2026 will showcase Nvidia's expanding ambitions. The conference, themed "AI Together," is expected to feature announcements of AI-focused chips and the rumored N1X laptop SoC—marking Nvidia's entry into the laptop processor market. Budget PC makers are also anticipated to debut products competing directly with Apple's rumored MacBook Neo.
These developments underscore Nvidia's strategy to dominate AI infrastructure from data centers to desktops to consumer laptops, combining proprietary hardware with open-source models and enterprise software tools.
Acer announced the Nitro Blaze Link, a Linux-based handheld designed as a streaming device for PC games rather than local play. The device launches in Q4 2026.
Microsoft unveiled the Surface Laptop Ultra, a new high-end mobile workstation powered by Nvidia's RTX Spark chip that marks the company's most direct challenge to Apple's MacBook Pro lineup.
Fluence Energy's stock jumped more than 40% after the US energy storage company announced a collaboration with Siemens and Nvidia to design artificial intelligence data centers.
Nvidia announced the RTX Spark, an Arm-based chip family designed for Windows laptops and mini-PCs, marking the company's entry into the consumer PC processor market alongside Intel, AMD, and Qualcomm. The chip will power Microsoft's Surface Laptop Ultra, launching this fall.