NVIDIA introduced the DGX Spark desktop solution for AI development and prototyping, delivering up to 1000 TOPS (FP4) of AI performance a few months ago. The company has now unveiled the Jetson AGX Thor Developer Kit, which looks similar at first, but it’s designed for edge AI and robotics with I/Os for cameras and sensors.

Nvidia’s NeMo Retriever models and retrieval-augmented generation pipeline make quick work of ingesting PDFs and generating reports based on them. Chalk one up for the plan-reflect-refine architecture.

Nvidia is set to beat expectations, with strong revenue forecast of $48b for Q2 and expected growth through 2026, driven by demand for AI and new Blackwell architecture.

Dylan Patel, founder of SemiAnalysis, talks about the AI hardware landscape, GPT-5, business models, and the future of AI infrastructure with A16z Venture

The company has recently raised its Q4 FY’25 forecast, expecting revenue of $11.2 billion at the midpoint, an increase from its previous estimate of $10.7

Nvidia stock has rebounded from its lows earlier this year, and expectations are mighty high as earnings come into view. While competition is rising, Nvidia remains well positioned to capitalize on a number of emerging infrastructure opportunities.

Nvidia ( NVDA 0.94%) owns stock in publicly traded companies as part of its treasury strategy. Its $4.3 billion portfolio included six positions as of the June-ending quarter, but two accounted for 95% of invested assets: 91% was allocated to CoreWeave ( CRWV -3.17%) and 4% was to Arm Holdings ( ARM 1.52%).