*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Software Techniques to Mitigate Errors in Noisy Quantum Computers
Committee:
Dr. Moinuddin Qurreshi, ECE, Chair , Advisor
Dr. Tushar Krishna, ECE
Dr. Vivek Sarkar, CoC
Dr. Asif Khan, ECE
Dr. Kenneth Brown, Duke
Abstract: Quantum computers are domain-specific accelerators that can provide a large speedup for important problems. Quantum computers with few tens of qubits have already been demonstrated, and machines with 100+ qubits are expected soon. These machines face significant reliability and scalability challenges. Due to limited and unreliable qubits, these machines are operated in the Noisy Intermediate Scale Quantum (NISQ) mode of computing. The computation on a NISQ machine can produce incorrect output. Therefore, in the NISQ mode, a program is run thousands of times, and the output log is analyzed to infer the correct output. However, the error rates on current quantum hardware are such that the likelihood of obtaining the right answer is still quite small for NISQ machines, and this problem only becomes worse for programs with a large number of instructions. This dissertation shows how the reliability of near-term quantum computers can be improved by developing software techniques. Our first work (ASPLOS 2019) exploits the variability in the error rates of qubits to steer more operations towards qubits with lower error rates and avoid error-prone qubits. Our second work (MICRO 2019a) looks at executing different versions of the programs tuned to cause diverse mistakes so that the machine is less vulnerable to correlated errors, thereby making it easier to infer the correct answer. Our third work (MICRO 2019b) looks at exploiting the state-dependent bias in measurement errors (state 1 is more error-prone than state 0) and dynamically flips the state of the qubit to measure the stronger state. We perform our evaluations on real quantum machines from IBM and demonstrate significant improvement in the overall system reliability.
