IBM just unveiled a major breakthrough in quantum computing technology. They announced a completely new quantum architecture that has the potential to fundamentally reshape the quantum processing landscape. Led by IBM Fellow Matthias Steffen, the quantum processor technology team aims to optimize the use of qubits, potentially requiring several hundred physical qubits to create just ten logical qubits. Just how many physical qubits they’ll need remains to be figured out, but that’s what Steffen and his collaborators are figuring out.
The crown jewel of this new hardware is their 120-qubit processor, Nighthawk. On the connectivity side, this newfangled processor features vastly superior connectivity than IBM’s last Heron processor. Nighthawk enables IBM customers to easily run more complex quantum circuits. Now, they’re able to leverage 15 times the logical gates as they were able to using Heron. This new level of connectivity is anticipated to improve computational efficiency and unlock possibilities for more sophisticated quantum algorithms.
These advances in coherence times represent a great leap forward for the Heron chips. They went from around 150 microseconds to a staggering 250 microseconds improvement. The IBM team is well aware that while scaling is needed, making better coherence times for the underlying qubits will be essential for further advances.
“We’ve cracked the code to quantum error correction and it’s our plan to build the first large-scale, fault-tolerant quantum computer,” said Jay Gambetta, highlighting the ambitious trajectory of IBM’s quantum computing efforts.
IBM has previously stated their goal is to construct a large-scale, fault-tolerant quantum computer called Starling. He says they want to put it in customers’ hands by 2029. Starling will take advantage of a new type of error-correction scheme called quantum low-density parity check (qLDPC) codes. This radical new approach will require roughly one-tenth the qubits compared to conventional surface codes. This reduction represents a major computational breakthrough in the field of quantum error correction.
The extended-range couplers of the new architecture allow for greater distances between parent and daughter layers. These couplers provide an important connection between physically distant qubits on the same chip. This design seeks to continue to push the performance and reliability envelope for large-scale future quantum systems. Notably, substantial engineering hurdles remain. We need to address every piece of infrastructure, such as connectors and amplifiers, to truly maximize everything the Nighthawk processor has to offer.
Fast forward to 2027 and it will be IBM introducing the next Earth-shattering system. This new device, aptly named Cockatoo, will ring three Nighthawk modules together. This move represents yet another step forward in IBM’s long, sustained push to develop practical quantum computing technology.
Mark Horvath just knows that sooner or later it’s gonna work. He believes we’re a lot farther from that goal than most folks assume – underscoring the challenges of building practical, large-scale quantum systems.
As part of her PhD research, Steffen analyzed the benefits of using qLDPC codes. He continued his enthusiasm by explaining how these codes reduce the overhead introduced by non-quantum processors.