Next-Gen Superchip Successor Unveiled by Nvidia as Rubin’s Replacement for Blackwell

The tech world is abuzz with excitement as the CEO of Nvidia, Jensen Huang, unveiled a groundbreaking roadmap for ARM's Adonis Neoverse V4 processor. This cutting-edge processor is set to follow a one-year release cycle, aligning with the upcoming successor to the Grace Hopper superchip. According to Huang, the Rubin platform will supersede the Blackwell platform and will introduce a host of new features, including innovative GPUs and a new ARM-based CPU named Vera.

At the heart of the Grace superchip are 144 ARM Neoverse V2 CPU cores equipped with Scalable Vector Extensions (SVE2). Nvidia, a key player in the tech industry, has previously ventured into developing its own version of ARM cores known as Denver. As a leading customer for ARM, Nvidia is gearing up to leverage the next-generation V4 architecture, codenamed Adonis, in the development of Vera. This integration will be part of a Compute Sub-System (CSS) dubbed Vega.

Not stopping at just processor advancements, Nvidia's chip launches will also incorporate cutting-edge networking technologies such as NVLink 6, CX9 SuperNIC, and the X1600 converged InfiniBand/Ethernet switch. Huang emphasized the significance of these developments during his keynote address at the Computex 2024 trade show in Taipei, Taiwan, highlighting Nvidia's commitment to pushing the boundaries of technology and staying ahead of the competition.

Furthermore, Taiwanese manufacturers like Foxconn, Pegatron, and Wistron, renowned for producing Apple devices, are gearing up to integrate the Blackwell GPU and digital twin technology into their operations. This move signifies a significant shift towards adopting advanced technologies to enhance efficiency and productivity in their factories.

Chip designer MediaTek is also joining forces with Nvidia by incorporating the Nvidia TAO training and pretrained models into its AI development workflow for IoT devices. This collaboration aims to bring state-of-the-art AI capabilities, such as Metropolis and advanced visual perception, to billions of IoT far-edge devices, streamlining software development and paving the way for MediaTek's growth in edge IoT.