MicroCloud Hologram, a technology service provider, proposed a feasible path for achieving computational advantages in the short term, centered around three core aspects: the collaborative utilization of QPUs, error control, and algorithm optimization.
First, by combining multiple QPUs through circuit knitting technology, it is possible to overcome the limitations of a single QPU in terms of qubit count and computational capacity, enabling larger-scale quantum computing tasks. Circuit knitting technology can decompose complex quantum circuits into multiple sub-circuits, which are then distributed to different QPUs for parallel processing. The computational results from each QPU are subsequently integrated through specific interface technologies to form a complete solution. Second, the application of error suppression and mitigation techniques can reduce the impact of errors on computational results under existing hardware conditions. Error suppression reduces error occurrence by optimizing quantum gate operations and improving qubit coherence, while error mitigation corrects errors that have already occurred through classical post-processing using statistical methods, thereby enhancing the quality of the solution. Finally, focusing on heuristic quantum algorithm versions with asymptotic speedup enables efficient solutions to practical problems under the constraints of current quantum hardware. Although heuristic algorithms do not guarantee finding the optimal solution, they can find near-optimal solutions within a reasonable time frame, and their asymptotic speedup characteristic ensures that their advantages over classical algorithms become increasingly evident as problem scale grows, making them suitable for practical applications in the short term.
Also Read: CIO Influence Interview with Liav Caspi, Co-Founder & CTO at Legit Security
To achieve the aforementioned short-term goals and lay the foundation for long-term development, HOLO proposes the need to build a “quantum-centric supercomputing” architecture, the core of which lies in achieving seamless integration between quantum processors and classical processors. In traditional computing architectures, classical processors hold the dominant position, with quantum processors often serving as auxiliary tools, and there are clear boundaries in data interaction and task scheduling between the two, making it difficult to fully leverage the performance of quantum processors. In contrast, the quantum-centric supercomputing architecture redefines the relationship between the two, assigning quantum-computable portions to quantum processors and classical-computable portions to classical processors based on the characteristics of the computing tasks, thereby forming a collaborative working mode.
As quantum computing hardware continues to upgrade, the system’s complexity also increases accordingly, including qubit manipulation, the implementation of quantum error correction, multi-QPU collaboration, and so on; these complexities pose significant barriers to users in utilizing quantum computing technology. Future quantum computing software needs to possess higher levels of automation and intelligence, by optimizing algorithm design, developing efficient compilation tools, and building user-friendly interfaces, to hide the underlying technical details, allowing users to conveniently invoke quantum computing resources just like using classical computers without needing to deeply understand the principles of quantum computing. The “frictionless” user experience proposed by HOLO can greatly lower the application threshold for quantum computing, promote the popularization of quantum computing technology in more fields, and achieve “ubiquitous” application scenarios, such as financial institutions using quantum computing for risk analysis, medical institutions using quantum computing to accelerate drug development, and so on, enabling quantum computing to truly integrate into daily life and the development of various industries.
Catch more CIO Insights: The CIO as AI Ethics Architect: Building Trust In The Algorithmic Enterprise
[To share your insights with us, please write to psen@itechseries.com ]
