IBM Solved Quantum Computing’s Biggest Problem

IBM has announced plans to build the world’s first large-scale quantum computer, Starling, by 2028. It will be available to users online by 2029 and will include a modular network of chips housed in a new data center in Poughkeepsie, New York.

According to IBM Vice President Jay Gambetta, the company has already begun building the infrastructure. He claims that they have solved a key problem in the industry: quantum error correction. This will give IBM an advantage over competitors including Google, Amazon, and startups like QuEra and PsiQuantum.

Quantum computers use qubits that can exist in a superposition of states, but they are susceptible to errors due to external influences. Without error correction, such machines cannot perform complex algorithms, such as precise chemical simulations for the development of new materials and drugs.

Starling will be able to perform 100 million logical operations in a row with high accuracy, while today’s quantum computers are limited to thousands of operations. One of Starling’s main advantages will be the ability to diagnose errors in real time using algorithms running on FPGA chips.

But experts doubt that even Starling’s 200 logical qubits will be enough to solve practical problems — useful algorithms may require billions of operations. After 2029, IBM plans to create a more powerful Blue Jay computer with 2,000 logical qubits and a billion operations.

Leave a Comment