Google & IBM: New Age of Quantum Computing is About to Begin
The race to build the first workable quantum computer has come alive, with industry leaders Google and IBM both claiming that they can produce full-scale systems within five years.
Recent technical breakthroughs have revived confidence in what was once considered little more than fantasy.
"It doesn't feel like a dream anymore," says Jay Gambetta, who is heading up IBM's VP of Quantum.
"I really do feel like we've cracked the code and we'll be able to build this machine by the end of the decade."
Why developing quantum computers is so difficult
This renewed optimism amongst Big Tech's quantum computing teams comes in spite of the formidable challenges that lay ahead of them.
Scaling quantum systems from laboratory experiments to industrial applications has, until now, always eluded the world's largest tech companies.
Current systems comprise fewer than 200 qubits, but commercial viability requires expanding to one million qubits or more.
Julian Kelly, Head of Hardware at Google Quantum AI, believes that his team can achieve this goal, though.
"All the engineering and scientific challenges are surmountable," he says.
Others, however, remain more cautious about timelines.
Oskar Painter, Director of Quantum Hardware at AWS, thinks that truly workable quantum computers are still 15-30 years away.
"The remaining hurdles seem technically less challenging than the fundamental physics, but we should not underestimate that engineering effort to scale," he explains.
Differences in approaches across the industry
The companies have pursued different strategies for overcoming quantum computing's inherent instability problem.
Google demonstrated quantum error correction that improves as system size increases, using a technique called surface code that connects each qubit to its nearest neighbours in a two-dimensional grid.
"Any company trying to scale up without first reaching this point would end up with a very expensive machine that outputs noise, and consumes power and a lot of people's time and engineering effort and does not provide any value at all," says Julian.
IBM has chosen a different path with low-density parity-check code, claiming it requires 90% fewer qubits than Google's approach.
However, this method depends on longer connections between distant qubits, which can make things more complex.
The problems facing engineers
Beyond the theoretical challenges, companies face numerous practical engineering problems.
These include redesigning the complex wiring systems found in current quantum computers and developing much larger specialised refrigeration units.
Quantum systems using superconductors must operate at temperatures close to absolute zero, requiring substantial infrastructure investments.
Google aims to reduce component costs by a factor of ten to achieve its target cost of US$1bn for a complete system.
IBM's experimental Condor chip, with 433 qubits, demonstrated the scaling problems when increased qubit numbers led to interference between components.
"Stacking larger numbers of qubits together like this creates a bizarre effect we can't control anymore," says Subodh Kulkarni, CEO of Rigetti Computing.
"That's a nasty physics problem to solve."
Government interest intensifies quantum competition
Despite the difficulties in developing quantum computing, the official recognition of the technology's potential has grown a great deal recently .
Darpa, the Pentagon's advanced research agency, launched a comprehensive study last year to identify which quantum companies could reach practical scale fastest.
The race has also attracted companies pursuing alternative approaches, including Amazon and Microsoft, which claim breakthrough qubit designs using exotic states of matter.
Despite the challenges, industry confidence remains high that practical quantum computers will emerge this decade, potentially revolutionising fields from materials science to AI.
