Details
- NVIDIA and Intel have announced a major multi-year partnership to co-develop custom x86 processors, as NVIDIA acquires roughly 5% of Intel through a $5 billion stock purchase at $23.28 per share.
- The collaboration will yield Intel-built x86 RTX system-on-chips (SOCs) featuring integrated NVIDIA RTX GPU chiplets for consumer PCs, alongside NVIDIA-customized x86 data center CPUs designed for advanced AI infrastructure.
- These new products will replace traditional PCIe links with NVIDIA's NVLink interface, enabling up to 14 times higher bandwidth, lower latency, and unified CPU-GPU memory access for greater performance.
- The effort builds on Intel’s earlier Kaby Lake-G project with AMD, but promises deeper integration and a unified driver approach to mitigate past support problems.
- Consumer-focused RTX SOCs aim at the thin-and-light gaming laptop and compact PC market, directly challenging AMD’s APUs; meanwhile, the custom x86 server chips target hyperscale and enterprise AI workloads.
Impact
This strategic alliance redefines the x86 landscape, with NVIDIA and Intel joining forces to counter AMD’s advances in both gaming and data center sectors. The $5 billion investment underscores Intel’s financial urgency but also signals commitment to the AI boom. As AMD pursues its own proprietary interconnects, this partnership could significantly shift momentum in the battles for gaming laptops and AI server dominance.