One of the biggest challenges in quaum computing is to reach steady state. But Google researchers have recely discovered a new technique using artificial ielligence to make the use of quaum computing practical in real life.
In a research paper published in the journal Nature, researchers from Google’s DeepMind unit explain that their new AI system, called AlphaQubit, has been remarkably successful in correcting errors that have long plagued quaum computers. This success could eveually pave the way for the use of quaum computing in everyday life.
Google researchers have discovered a new way to correct quaum errors
For practical use, quaum computers require an error rate of only one per trillion operations (10-^12), but curre hardware has an error rate of between 10-3 and 10-2 per operation, making error correction critical. In a stateme published by Google, it is stated:
“Quaum computers have the poteial to revolutionize drug discovery, materials design, and fundameal physics, provided we can get them to work reliably. But nothing is perfect. Quaum systems are incredibly fragile. “Even the slightest environmeal ierference from heat, vibration, electromagnetic fields or even cosmic rays can disrupt their delicate quaum states and lead to errors that make calculations unreliable.”
Google also we on to say that quaum computers can solve problems in just a few hours that would take ordinary computers billions of years to solve, but the new processors are more susceptible to noise than conveional processors.
Google’s AlphaQubit artificial ielligence system aims to tackle this problem. This system uses a complex neural network architecture that has shown unprecedeed accuracy in detecting and correcting quaum errors. In fact, AlphaQubit shows about 6% less error than the best previous methods in large experimes and 30% less error than traditional techniques.
Although AlphaQubit is considered a step forward and has excelle performance in accurately detecting faults, this system is still very slow in correcting faults in a superconducting processor in real time.
