Quantum Circuit Partitioning for Storage-Based Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quantum circuit simulation technologies require exponentially increasing memory and are costly, making it difficult to simulate circuits larger than 50 qubits using today's supercomputers, which are inaccessible to ordinary users.
Innovation Solution
A quantum circuit simulation system that partitions quantum circuits into sub-circuits and uses a storage system with cheaper, high-capacity devices like HDDs, SSDs, or NVMe, optimizing memory access through permutation operations and prefetching techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If quantum circuit simulation is performed using conventional memory, then simulation accuracy is maintained, but memory cost and requirements increase exponentially
Solution Approach 1:
The patent divides the quantum circuit into multiple sub-circuits and processes them sequentially. Instead of loading the entire quantum circuit state into memory at once, the system processes sub-circuits one by one, significantly reducing the peak memory requirement from exponential (2^N) to manageable levels while maintaining simulation accuracy through proper state management between sub-circuit executions.
Solution Approach 2:
The patent replaces expensive, high-speed memory (DRAM) with cheaper storage devices (HDD, SSD, NVMe) for storing quantum circuit data. The system accepts that storage access is slower but compensates through optimized access patterns and prefetching, achieving cost-effective simulation by using inexpensive storage media instead of costly memory.
2Productivity
If supercomputers are used to handle enormous memory requirements, then simulation capability is improved, but cost and accessibility worsen
Solution Approach 1:
The patent demonstrates that quantum circuit simulation can be performed on ordinary computers equipped with conventional storage devices (HDD, SSD, NVMe) instead of requiring expensive supercomputers. By optimizing the simulation approach to work with limited memory and using cheaper storage media, the system makes quantum circuit simulation accessible to ordinary users and organizations without supercomputer budgets.
Solution Approach 2:
By segmenting the quantum circuit into sub-circuits and processing them sequentially with optimized memory management, the patent enables simulation on commodity hardware. This approach breaks down the computationally intensive task into manageable chunks that can be handled by ordinary computers, eliminating the need for supercomputing resources.
3Quantity of substance
If storage devices are used instead of memory, then storage capacity and cost are improved, but access speed deteriorates
Solution Approach 1:
The patent implements a prefetching mechanism that loads data from storage devices into memory in advance before it is actually needed for processing. By anticipating future data access requirements and pre-loading the data, the system mitigates the slow access speed of storage devices, ensuring that data is ready when needed without causing processing delays.
Solution Approach 2:
The patent optimizes the simulation workflow to maintain continuous processing by coordinating data transfer from storage with computation operations. Through persistent prefetching and optimized access patterns, the system ensures that storage operations do not create idle periods in the processing pipeline, keeping the system continuously productive despite using slower storage media.
Data Source
AI summary
A quantum circuit simulation device according to the present disclosure includes a storage system including one or more storage devices, and a host system connected to the storage system and configured to simulate a quantum circuit, wherein the quantum circuit stored in the storage system may be partitioned into one or more sub-circuits, and the host system may sequentially simulate the one or more sub-circuits in a main memory.