IBM is ramping up its production of the Quantum Loon chip, a key component in its quantum computing efforts. This follows through on promises made by IBM back in June. The new architecture represents two significant changes for the company: it's shifting to nearest-neighbor connections and adding long-distance connections. These changes aim to increase efficiency in error-corrected logical qubits.
IBM has built two processors, one called Loon and another called Nighthawk, both of which use a square grid structure with nearest-neighbor connections but lack the long-distance connections found in Loon. The focus with Nighthawk is on reducing error rates to allow researchers to test quantum algorithms for advantage. A GitHub repository will be launched to deposit code and performance data for classical and quantum algorithms.
Meanwhile, Oxford Ionics has achieved record-low error rates in its qubit operations using trapped ions. However, the company was ultimately acquired by a competitor, meaning it's no longer making major announcements about its technology.
IonQ is now building on what it gained from Oxford Ionics, announcing an even lower error rate for two-qubit gates and describing a method that doesn't require full cooling of the ions. This could allow the entire machine to operate with its ions at a slightly elevated temperature, avoiding one of the two cooling steps.
Quantum Art has announced a collaboration with Nvidia on a more efficient compiler for operations on its hardware. The company's approach to boosting efficiency involves performing gates with much larger collections of ions and using lasers to "pin" them in place.
IBM has built two processors, one called Loon and another called Nighthawk, both of which use a square grid structure with nearest-neighbor connections but lack the long-distance connections found in Loon. The focus with Nighthawk is on reducing error rates to allow researchers to test quantum algorithms for advantage. A GitHub repository will be launched to deposit code and performance data for classical and quantum algorithms.
Meanwhile, Oxford Ionics has achieved record-low error rates in its qubit operations using trapped ions. However, the company was ultimately acquired by a competitor, meaning it's no longer making major announcements about its technology.
IonQ is now building on what it gained from Oxford Ionics, announcing an even lower error rate for two-qubit gates and describing a method that doesn't require full cooling of the ions. This could allow the entire machine to operate with its ions at a slightly elevated temperature, avoiding one of the two cooling steps.
Quantum Art has announced a collaboration with Nvidia on a more efficient compiler for operations on its hardware. The company's approach to boosting efficiency involves performing gates with much larger collections of ions and using lasers to "pin" them in place.