Today, we're witnessing a thrilling trend with the rapid expansion and availability of AI-based applications across various edge devices. As AI progresses, providing AI researchers, data scientists, developers, and students access to high-performance computing is essential for developing or running advanced models, such as those for language, vision, or multi-modal applications. The swift pace of AI innovation demands broad access to performance capabilities beyond the cloud, directly at the edge and in developers' hands. NVIDIA’s breakthrough toward this vision is Project DIGITS, a personal AI supercomputer announced by NVIDIA’s Founder and CEO Jensen Huang at CES 2025. This Linux-based system, featuring Arm-powered CPU cores, offers every AI developer a high-performance AI system right on their desk. Powered by the NVIDIA GB10 Grace Blackwell Superchip, it combines the NVIDIA Grace CPU and Blackwell GPU with the latest CUDA cores, RT Cores, and 128GB of unified memory via the NVIDIA NVLink interconnect. The setup boasts a remarkable petaflop performance at FP4 precision, enabling developers to prototype, fine-tune, and run inferencing with large AI models jointly with cloud resources or data centers. Utilizing the Arm compute platform allows new AI models to perform efficiently at the edge. This platform is central to today's edge devices and aims at high-performance market entries, such as in Project DIGITS.

These technologies are tailored for optimal performance and efficiency across applications. In Project DIGITS, the Arm-based CPU and GPU work together to cover varied workloads, essential for peak AI performance while managing memory use and power consumption. Ashish Karandikar, VP of SoC Products at NVIDIA, emphasizes that the collaboration with Arm on the GB10 Superchip accelerates AI innovation by merging NVIDIA’s AI skills with Arm's scalable compute platform, now making powerful computing accessible to every AI developer and researcher. For developers, having an entirely integrated hardware and software AI platform is crucial. Project DIGITS utilizes the Linux system and an expansive NVIDIA AI software library, which supports developers in advancing their AI workflows. The common platform architecture between Arm data centers and edge environments allows developers to seamlessly use consistent tools for AI application creation. With over 20 million developers building on Arm's platform, the open-source community is innovating rapidly. As the world's leading compute platform for AI, Arm’s suitability for the GB10 Superchip used in Project DIGITS allows it to handle AI models up to 200B parameters, expanding capabilities for millions of developers worldwide. In collaboration with NVIDIA, this lays the groundwork for a new wave of revolutionary AI applications.