secunet, Red Hat and deepset are building an AI runtime service for German authorities handling classified information, built on Red Hat AI Factory with NVIDIA.
SpaceX disclosed a binding commitment for an Apollo-managed entity to buy GPUs and lease them back, with about $40 billion in principal payments over roughly five years.
SpaceX’s $40 billion debt raise, led by Apollo and Goldman Sachs, is being finalized at about 8.3% on its riskiest slice and around 7.75% on a chip-backed tranche, Semafor reports.
Today’s roundup: SpaceX, Tesla, Red Hat, SUSE news from the past day, with links to full write-ups.
SUSE and Fujitsu (FSAS) are combining SUSE AI Factory with NVIDIA and NVIDIA CUDA-Q on Fujitsu servers for on-premises hybrid quantum workloads, due in late 2026.
SUSE announced SLES for NVIDIA, built on SLES 16, and NVIDIA OpenShell integration for SUSE AI Factory with NVIDIA. General availability is planned for November 2026.
At NVIDIA GTC Berlin, Red Hat added NVIDIA Vera CPU support to RHEL for NVIDIA, announced Day 0 Vera Rubin enablement and a joint vLLM/llm-d inference reference design.
Linus Torvalds’ famous one-finger message to Nvidia came from a June 2012 Q&A at Aalto University in Finland. Here is the context and what has changed since.
