Qedma, a developer of quantum noise resilience solutions, announced on Thursday that it has raised $26 million in Series A funding. The round was led by Glilot Capital Partners through its early growth fund, Glilot+, with new participation from IBM, Korean Investment Partners, and others, alongside existing investors including TPY Capital. Qedma’s software solution addresses the critical challenge of reducing errors in quantum computing.
Quantum computers, by their nature, are extremely susceptible to errors from environmental “noise.” Unlike classical bits, which flip between zero and one, quantum bits, or qubits, can exist in a fragile superposition of states, making them vulnerable to the slightest disturbance. In theory, layers of error correction can stabilize these qubits, but the cost is enormous: today’s error-correction schemes require as many as 1,000 physical qubits to reliably preserve a single logical one.
Qedma’s founders say their approach can ease this burden dramatically. By analyzing and learning the unique noise profile of each quantum device, Qedma’s software adjusts the algorithms to suppress certain types of errors before they happen and mitigate others through post-processing. The company claims its method can enable quantum calculations up to 1,000 times larger than what today’s hardware alone allows.
The idea traces back to a chance lunch in 2020 between Prof. Netanel Lindner (CTO), a quantum physicist at the Technion, and Dr. Asif Sinay (CEO), a physicist turned tech executive. Lindner shared his view that device-specific noise understanding might hold the key. Independently, Prof. Dorit Aharonov (Chief Scientific Officer), a pioneer behind the fault tolerance theorem that proved quantum error correction was theoretically possible, shared a similar vision with Sinay. Weekly conversations between the three turned into a startup that now sees itself as building the “operating layer” quantum hardware desperately needs.
“In every algorithm there are lines of code where each operation has a reliability of 99%, which creates a problem because the result may be incorrect,” Sinay told Calcalist. “When you connect a large number of arithmetic operations, hundreds of thousands of logical operations, the chance of errors becomes very high due to the nature of quantum computers. Unlike classical computing, you cannot simply stop and correct mid-process. We have developed a method to reduce these errors and perform calculations that classical computing cannot handle. We know how to run algorithms with 99% accuracy. This is the critical part of quantum computing: without an operating layer that corrects errors, we’re just selling white elephants. We have about 40 employees who focus on the software layer and on removing errors, because without that, you can’t run any computing at scale.”
He added, “We have a product and we’ve integrated with IBM’s quantum computers. Every customer who accesses IBM’s hardware can also access our software, and we’re integrating with other hardware companies in our field,” Sinay said. “The entire industry knows that without error correction, quantum computing is useless. Qubits must be as noise-free as possible. In this round, IBM is investing for the first time in a company that’s building what’s closest to an operating system for quantum computing. IBM did the most rigorous due diligence on us and told us it believes in us, investing millions of dollars, not only for a stake in the company, but also to help advance it.”
“Lior Litwak from the Glilot fund, who was in my Talpiot class, decided we were the best investment in quantum computing. We also wanted an East Asian partner, which is why we brought in the Korean fund. Every country today must have quantum computing capability,” Sinay said.