Researchers on the College of Chicago and Argonne Nationwide Laboratory considerably diminished this hole through the use of information compression methods to suit a 61-qubit simulation of Grover’s quantum search algorithm on a big supercomputer with zero.four p.c error. Different quantum algorithms have been additionally simulated with considerably extra qubits and quantum gates than earlier efforts.

Classical simulation of quantum circuits is essential for higher understanding the operations and behaviors of quantum computation. Nevertheless, at the moment’s sensible full-state simulation restrict is 48 qubits, as a result of the variety of quantum state amplitudes required for the total simulation will increase exponentially with the variety of qubits, making bodily reminiscence the limiting issue. Given n qubits, scientists want 2^n amplitudes to explain the quantum system.

There are already a number of present methods that commerce execution time for reminiscence area. For various functions, individuals select completely different simulation methods. This work supplies yet one more possibility within the set of instruments to scale quantum circuit simulation, making use of lossless and lossy information compression methods to the state vectors.

Determine 1 reveals an outline of our simulation design. The Message Passing Interface (MPI) is used to execute the simulation in parallel. Assuming we simulate n-qubit methods and have r ranks in whole, the state vector is split equally on r ranks, and every partial state vector is split into nb blocks on every rank. Every block is saved in a compressed format within the reminiscence.

Learn extra from College of Chicago’s Division of Laptop Science.


Please enter your comment!
Please enter your name here