Smaller, faster, cheaper: QuamCore rethinks how to build superconducting quantum computers

A newly funded start-up claims to have the key to helping some quantum computing firms adopt a faster and more financially feasible approach for achieving maximum scalability, practicality, and commercial viability.

QuamCore, a three-year-old company based in Herzliya, Israel, recently emerged from stealth mode with $9 million in seed funding led by Viola Ventures and Earth & Beyond Ventures.

The company has developed a new superconducting quantum processor architecture that it says will enable the integration of 1 million qubits into a single cryostat, the super-cold refrigerator in which quantum computers built on superconducting qubits do much of their magic.

Superconducting qubits rely on electrical circuits to create artificial atoms for qubit production. Three giants of the quantum computing space–IBM, Google, and Microsoft–are currently developing quantum computing machines and roadmaps based on superconducting technology, which promises to eventually have faster gate-model operation than other techniques and other benefits, but typically requires a long and costly path to achieve scalability, due to needs for complex, expensive analog electronics and cabling, large physical spaces, massive power consumption, and more supporting equipment. Current superconducting systems remain in the range of thousands of qubits for now, and even in these cases only a small number of firms with very deep pockets (the above-named big three) have the resources to build them.

A QuamCore press release noted, “Traditional control electronics generate too much heat and must therefore be placed outside the cryostat… This separation creates an insurmountable scaling bottleneck, requiring millions of cables to connect the control electronics to the processor. Right now, the most advanced quantum computers from IBM and Google can only fit about 5,000 qubits per cryostat, requiring hundreds of interconnected cryostats in a football-field-sized facility to scale.”

QuamCore CEO Dr. Alon Cohen, former head of Intel Mobileye’s EyeC Radar Group, told Fierce Electronics via e-mail that the company, by rethinking the superconducting processor architecture and introducing digital control capabilities, can reduce the size, energy consumption, and cost of such systems.

“Moving to digital superconducting logic allowed us to reduce the controller’s power consumption by a factor of 1 billion,” he said. “This massive reduction overcame the previous limitation that required qubit control from room temperature, making it feasible to control qubits directly inside the cryogenic environment. While analog components such as mixers, ADCs, DACs, and amplifiers typically consume orders of magnitude more power, our transition from CMOS to superconducting digital logic yielded an additional reduction of several orders of magnitude. These combined improvements are essential for controlling the qubits within a 10 mK [Absolute Zero, close to -460 degrees Fahrenheit] cryostat without generating too much heat to disrupt their fragile quantum states.”

Cohen added that QuamCore is not looking to disrupt the ongoing superconducting work of the likes of IBM and Google, but complement those efforts. “By enabling more qubits to fit within a single cryostat, our solution allows computational scaling (more qubits) without an equivalent physical expansion of hardware. We view this as accelerating the progress already being made by the industry’s leading innovators. Our technology is compatible with all types of superconducting qubits.”

Though QuamCore believes it can help a larger number of companies to build more scalable superconducting quantum computers more feasibly and faster than current roadmaps call for, Cohen admitted the company still has a long journey ahead of it.

“Building on these advancements, we plan to produce our first “small” quantum computer based entirely on this technology within two years,” he said. “This smaller-scale system will demonstrate the method and validate its feasibility. According to our roadmap, we expect to reach 1 million qubits by 2033–2034, paving the way for truly large-scale quantum computing and start unlocking real world applications.”