- Tim Costa, Senior Director of quantum computing at Nvidia, explains his company’s role in quantum computing as providing GPU-accelerated tools, software, and infrastructure to support quantum companies in developing QPUs, integrating them into data centers alongside CPUs and GPUs, without building quantum computers themselves.
- Nvidia collaborates with over 160 quantum groups, using AI-driven supercomputers for simulations, error correction, and interconnect development, while exploring how quantum outputs could train AI models, accelerating the timeline to useful quantum computing.
- Costa highlighted Nvidia’s upcoming Boston research center as a key step to physically connect quantum devices with GPUs and predicted chemistry and biochemistry as the first fields quantum computing will disrupt due to their alignment with quantum physics.
During a special edition of “Bloomberg Technology” from Nvidia’s GTC, Tim Costa, Senior Director of quantum computing at Nvidia (NVDA), joined host Ed Ludlow to discuss Nvidia’s role in advancing quantum computing without building or selling quantum computers themselves. Costa explained that Nvidia envisions quantum processors integrating into large-scale data centers alongside CPUs, GPUs, and other components, forming a heterogeneous computing ecosystem where each part plays to its strengths, with quantum technology excelling at specific computations. Nvidia supports this vision by aiding quantum companies in developing better quantum processing units (QPUs) through GPU-accelerated simulations, algorithm design, and infrastructure like error correction and device calibration, leveraging AI-driven supercomputers to manage the complexities of these physics-based systems.
Costa highlighted Nvidia’s extensive collaboration with over 160 quantum computing groups, including partners showcased at GTC, who use Nvidia’s GPUs, software stacks, and interconnect technologies to refine QPUs and pursue scalable error correction using AI methods. He addressed the dual relationship between AI and quantum computing, noting that AI not only accelerates quantum device control and error correction – shortening the timeline to useful quantum computing – but also that quantum computers, as physics experiments, could generate data to train AI models, potentially unlocking answers to long-standing questions in quantum physics. This interplay was exemplified by discussions with Nvidia CEO Jensen Huang, where quantum outputs were proposed as inputs for training foundation models, showcasing a symbiotic potential between the technologies.
Looking ahead, Costa emphasized the significance of Nvidia’s new research center in Boston, a physical space where Nvidia and its partners can connect quantum devices to GPUs, develop interconnects, and pioneer error correction technologies, laying the groundwork for quantum-accelerated supercomputers. He predicted that chemistry and biochemistry would be the first fields disrupted by quantum computing, given the natural alignment between quantum physics in QPUs and the demands of accurate chemical modeling, a view he believes holds promise despite varied opinions within the industry. Throughout the conversation, Costa underscored Nvidia’s role in accelerating quantum progress, not by creating quantum hardware, but by providing the computational backbone and collaborative framework to bring this new era of computing to fruition.
WallStreetPit does not provide investment advice. All rights reserved.
Leave a Reply