Intel will build custom x86 CPUs for Nvidia's AI infrastructure as world's largest company invests $5 billion in beleaguered tech firm - and don't discount a data center x86 APU (www.techradar.com)

Intel and Nvidia announced a surprise strategic partnership in which Nvidia will invest $5 billion in Intel common stock and Intel will design custom x86 CPUs for Nvidia’s data‑center platforms while also building new x86 system‑on‑chips that integrate Nvidia RTX GPU chiplets for PCs. The companies said they will develop multiple generations of products: Nvidia‑branded data‑center servers will use Intel‑customized x86 processors, and consumer devices (gaming laptops and compact PCs) will ship Intel x86 “RTX SoCs” that pair Intel CPU cores with Nvidia GPU chiplets connected via NVLink. Timelines are early-stage and regulatory approval and manufacturing details (Intel-only fabs vs. TSMC support) remain unconfirmed. This deal is significant because it tightly couples Nvidia’s CUDA‑centric accelerated computing stack with the vast x86 ecosystem, potentially reshaping server and client architectures for AI workloads. For AI/ML practitioners, custom x86 CPUs tailored to Nvidia’s infrastructure could lower host–GPU bottlenecks, enable tighter co‑design across silicon and software, and create new APU‑style configurations for data centers. It also advances Intel’s IDM 2.0 strategy and gives Nvidia an alternative to, not a replacement for, its Arm‑based Grace line. Key open questions include product timelines, packaging and process node choices, and how this will affect competition with AMD, Arm‑based vendors, and Nvidia’s own CPU roadmap.