Nvidia has invested $2 billion in Marvell Technology to develop custom XPU chips integrated with Nvidia's NVLink technology. This partnership aims to enhance data center processing capabilities by combining Marvell's chip design expertise with Nvidia's high-speed interconnects. The collaboration targets accelerating AI workloads and improving system performance in cloud and edge computing environments. Nvidia and Marvell plan to release the first integrated solutions by late 2024, with broader deployment expected in 2025.
Source: source. Read the original story →