Abstract

Quantum computing has long been hailed as the next technological frontier. However, it has become increasingly clear that realizing the potential of Quantum Computers would require careful architecture and system design. One of the most critical elements, particularly limiting scale, is the classical control system. Developing thousands of highly synchronized channels with specifically well-designed analog specifications, combined with tight integration of real-time and near-real-time classical processing for quantum control pulse generation and readout signal processing, is crucial. Moreover, programming languages, tools, and environments that emphasize flexibility and ease-of-use are essential to maximize productivity in research and development. This keynote session will present the considerations for engineering control platforms that allow scaling up to 1,000 qubits and beyond – while maintaining the required performance, flexibility, and development productivity critical to the progress of the field. During the session, we will discuss the importance of defining quantum-classical processing requirements and benchmarks. We will also explore what to expect from quantum computers entering HPC centers and elaborate on NVIDIA-Quantum Machine’s DGX Quantum, an architecture designed to scale up ultra-low latency quantum-classical machines for practical implementations within supercomputers. Throughout, this talk will focus on practical solutions and essential cross-industry collaboration, defusing the hype and proposing a clear roadmap for scaling quantum control to accommodate hundreds and thousands of qubits.