SpaceX Plows $40B Into Nvidia Chips For AI Cloud
SpaceX reached a Binding Commitment for a $40 billion master finance lease with an entity managed by Apollo Global Management to buy NVIDIA Vera Rubin GPUs
Coverage of the chips that train and run AI, including GPUs, custom accelerators, supply constraints and the companies competing to build them.
SpaceX reached a Binding Commitment for a $40 billion master finance lease with an entity managed by Apollo Global Management to buy NVIDIA Vera Rubin GPUs
Zagitta released an open-source GPU unlock mod that lifts FP16 and BF16 issue-rate limits on some GeForce RTX 20, 30 and 40 cards, boosting tensor throughput
Synopsys said Nvidia Senior Manager for Hardware Engineering Ramulu Undevalli described his team’s use of its IC Validator software for physical
Kazakhtelecom deployed 400 NVIDIA H200 and 480 B300 GPUs in Astana and Almaty, adding 880 accelerators to train Kazakh-language models and run low-latency
Infinity hired a former Meta Superintelligence Labs chip veteran and the CUDA educator who has taught 2 million engineers to cut model and kernel porting
Blackwell architecture may raise AI throughput, but fab capacity, packaging, export controls and hyperscaler capex still limit frontier LLM scale.
GlobalFoundries will operate two new semiconductor factories in upstate New York backed by $21 billion the Commerce Department tied to a U.S.-Japan trade
A memory chip squeeze could slow AI, even after OpenAI’s $250 billion cloud deal, because only a few suppliers make the memory every AI processor needs.
Denser GPU. Vultr upgraded its cloud to serve agentic AI, adding the latest generation NVIDIA and AMD GPUs and platform upgrades for enterprise agentic AI.
Nebius expanded its AI Builder Program to Europe with $10 million in compute grants, giving early-stage AI teams a cheaper path to train and ship models.