May 23, 2019 — IonQ, a leader in trapped-ion quantum computing, has released the results of two rigorous real-world tests that show that its quantum computer can solve significantly more complex problems with greater accuracy than results published for any other quantum computer. The results, released on arXiv.org, are further validation of IonQ’s trapped-ion technology. In December, IonQ announced the most powerful system built to date. And in February, it published the world’s first quantum computer simulation of the water molecule.
“There is a lot of hype in the market now as many players release highly selective descriptions and statistics for their quantum computers,” said Christopher Monroe, IonQ’s co-founder and Chief Executive Officer.
“The real test of any computer is what can it do in a real-world setting. We challenged our machine with tough versions of two well-known algorithms that demonstrate the advantages of quantum computing over conventional devices. The IonQ quantum computer proved it could handle them. Practical benchmarks like these are what we need to see throughout the industry.”
The benchmarks released today, using the Bernstein-Vazirani algorithm and the Hidden Shift algorithm, each ask a computer to find a hidden number out of a set of possible numbers. For each algorithm, IonQ has solved for all possible 1024 hidden numbers with their 11-qubit machine, a more difficult task than has been attempted before on a quantum computer. Each test is based on a formula that a conventional computer can only solve by brute force, checking each potential result until it finds the answer. Quantum computers, in theory, can find these solutions in a single calculation by measuring all possible outputs simultaneously.
In its test of the Bernstein-Vazirani algorithm, IonQ’s computer tried to solve each of the 1024 variations 500 times. In total, it identified the correct answer in a single attempt 78% of the time. For the more difficult Hidden Shift algorithm, the IonQ machine found the correct answer in a single attempt in 35% of its trials. The lower accuracy is to be expected because the Hidden Shift algorithm requires between 35 and 50 quantum logic gates, including 10 more error-prone two-qubit gates. By comparison, conventional computers only get the correct answer to the Bernstein-Vazirani and Hidden Shift problems 0.1% of the time in a single attempt.
“In test after test, we are showing the power of IonQ’s trapped-ion technology to solve meaningful problems,” said Jungsang Kim, IonQ’s co-founder and Chief Strategy Officer. “Our qubits are perfect, and we know how to continually increase computing capacity through improvements in control engineering. Our qubits are also wired in software, allowing reconfigurable connections that can efficiently map onto any future problem.”