Welcome to Weijie Semiconductor

The synergy between QPU and GPU, classical supercomputing, and quantum computing

Tuesday, May 21, 2024

Quantum computing QPU

On ISC, Nvidia stated that it is accelerating quantum computing research and will connect the CUDA-Q quantum computing platform to supercomputer sites in Germany, Japan, and Poland, providing strong power for quantum processing units (QPUs) within its high-performance computer systems.
The QPU quantum processing unit is the core of quantum computers, which can utilize the behavior of particles such as electrons or photons to perform specific types of calculations, and its speed is much faster than any processor in today's computers. QPU's quantum computing is a computing technique based on the principles of quantum mechanics, utilizing quantum physical properties such as quantum superposition and entanglement to perform calculations. With these properties, QPU can quickly process and manipulate quantum states, completing complex calculations.
According to Nvidia, "holding a quantum processing unit (QPU) in hand may look and feel very similar to a graphics or data processing unit (DPU). They are very typical chips or modules with multiple chips. However, once run, QPU will explode with completely different forces.".
The performance of QPU is generally represented by the number of quantum bits it contains, and quantum bits are an abstract concept, so various techniques are needed to simulate this quantum bit node. There are currently many technological means in the manufacturing of qubits, such as superconducting qubits, ion trap qubits, quantum dot qubits, etc.
For example, this year's fault-tolerant quantum computer hardware developer Alice&Bob successfully chip a 16 qubit QPU - Helium 1, and Intel's 12 qubit quantum processing unit QPU "Tunnel Waterfall" from last year is also Intel's most advanced silicon spin quantum bit chip to date.
Quantum computing is still in its early stages, and it is unclear which technology will widely use quantum bits in QPU in the future. But what can be lacking is that theoretically, QPU requires less power and generates less heat than these classic processors today.

The Collaborative Work of Classical Supercomputing and Quantum Computing

The industry believes that with the development of QPU, there will be many collaborative scenarios of classical and quantum computing in the future, and computing tasks need to be able to run well on QPU, CPU, and GPU.
The NvidiaCUDA-Q quantum computing platform program can run on the QPU of quantum computers and GPUs simulating QPU in classical systems. ABCI-Q supercomputers from the Japan Institute of Industry and Technology, JSC from Germany, and PSNC from Poland, all of which are connected to the NvidiaCUDA-Q quantum computing platform on ISC, are conducting research on the hybrid of classical and quantum computing.
The Japan Institute of Industry and Technology has stated that ABCI-Q will make progress in practical quantum computing applications by utilizing a hybrid of quantum and classical acceleration computers; The Polish Poznan Supercomputing and Networking Center states that by constructing a new hybrid system of quantum and classical, efficient management, open and easy integration and programming can be achieved on future multi QPU and GPU systems; The German Ulrich Supercomputing Center also stated that hybrid quantum and classical accelerated supercomputing make quantum computing closer to reality.
A hybrid quantum system will undoubtedly bring quantum computing closer to reality and practical applications, and this hybrid system will solve complex problems that cannot be solved solely by classical computing in the future.

Summary

As an early stage technology track, every breakthrough in new technologies and hardware is pushing the quantum computing industry closer to reality. Ultimately, the disruptive innovation brought together by each breakthrough will quickly cause industry change in a short period of time. This technology that can change the world is moving step by step towards surpassing traditional classical computing.

Leave your comment