Jensen Huang, the co-founder and CEO of Nvidia, recently made some remarks that have stirred quite a reaction. “If you kind of said 15 years … that’d probably be on the early side. If you said 30 years, it’s probably on the late side. But if you picked 20 years, I think a whole bunch of us would believe it,” he said during an investor conference. Interestingly, Huang’s comments have led to a massive drop in the shares of several companies involved in the development of quantum computers.
Huang isn’t alone in his assessment. Many experts agree that fully functional quantum computers are still at least two decades away. Ivana Delevska is the investment chief of Spear Invest, which holds shares in quantum computing companies Rigetti and IonQ. She supports Huang’s view, “The 15 to 20-year timeline seems very realistic. That is roughly what it took Nvidia to develop accelerated computing.”
IBM and Google’s Roadmap: An Optimistic Outlook on the Future
It’s important to clarify what a fully functional quantum computer really means, or at least what Huang refers to as a “truly useful” quantum computer. If you view current quantum computers as prototypes, a fully functional machine would be one that can correct its own errors. This capability would allow users to solve a much broader range of problems than today’s prototypes can handle.
Some scientists estimate that achieving the long-awaited error correction will require hundreds of thousands of qubits. Meanwhile, others believe it may take several million qubits to reach this milestone. Regardless, fully functional quantum computers will only become a reality when the scientific community can develop quantum systems composed of a vast number of qubits. The challenge is that interconnecting and controlling a large number of qubits isn’t easy.
Companies will develop fully functional quantum computers when it becomes possible to create quantum systems comprised of a large number of qubits.
In practical terms, a quantum computer is a machine made up of several functional blocks, each containing interconnected qubits. This architecture is quite complex, and its inherent complexity makes scalability a significant challenge. However, this challenge hasn’t deterred companies like IBM, Google, and Microsoft from achieving important milestones, which allows for a reasonably optimistic outlook on the future.
If IBM adheres to its roadmap, it plans to release Starling, its first quantum computer equipped with the capability to correct its own errors, by 2029. The company is clearly aiming for error correction rather than mere error mitigation–an important distinction. The hardware should enable researchers to tackle problems that current prototypes can’t address. Additionally, by 2033, Blue Jay is expected to lead to the massive scaling of IBM’s error-correcting quantum hardware.
For its part, Google introduced Willow in early December. According to its creators, Willow is an innovative quantum processor that can exponentially reduce errors as the number of qubits increases. Historically, as the number of qubits in a quantum computer grew, managing noise and interactions between them became increasingly challenging. As such, error correction was difficult.
However, Google’s proposal would achieve what’s known as “being below threshold.” This ensures that the quantum computer won’t behave like a classical computer, preserving the advantages offered by its quantum nature.
Moreover, when Google is able to increase the number of qubits on its quantum chips significantly, its real-time error correction technology should enable the hardware to tackle a wide array of problems. If successful, this milestone could signal the arrival of fully functional quantum computers. It remains to be seen whether Huang’s predictions will hold true or not.
Image | IBM Research
View 0 comments