Finding out how to pack as many high-quality qubits as possible on a single quantum processor is a challenge that still keeps most researchers scratching their heads – but now quantum startup Rigetti Computing has come up with a radically new approach to the problem.
Instead of focusing on increasing the size of a single quantum processor, Rigetti has linked up various smaller chips together to create, instead, a modular processor that still has a higher overall qubit count.
Describing the technology as the world's "first multi-chip quantum processor", the company launched the device with the objective of reaching 80 qubits later this year, up from the current 31 qubits supported by its Aspen processor.
SEE: Building the bionic brain (free PDF) (TechRepublic)
By that time, the new quantum system will be available for Rigetti customers to use over the firm's Quantum Cloud Services platform.
"We've developed a fundamentally new approach to scaling quantum computers," said Chad Rigetti, the founder of Rigetti Computing. "Our proprietary innovations in chip design and manufacturing have unlocked what we believe is the fastest path to building the systems needed to run practical applications and error correction."
Like IBM and Google, Rigetti's quantum systems are based on superconducting qubits, which are mounted in arrays on a processor where they are coupled and controlled thanks to microwave pulses. Qubits are also connected to a resonator and associated wiring, which enables the system to encode, manipulate and read out quantum information.
Qubits come with special quantum properties that are expected to lend quantum computers unprecedented computational power. But for that to happen, processors will need to pack a significant number of qubits – far more than they currently do.
For quantum computers to start generating very early value, experts anticipate that at least 1,000 qubits will be necessary; and a million qubits is often cited as the threshold for most useful applications. In contrast, the most powerful quantum processors currently support less than 100 qubits.
Scaling up the number of qubits sitting on a single processor, however, is difficult. This is mostly due to the fragility of qubits, which need to be kept in ultra-protected environments that are colder than outer space to ensure that they remain in their quantum state. More qubits on a chip, therefore, inevitably mean more potential for failure and lower manufacturing yields.
Instead, Rigetti proposes to connect several identical processors, such as those that the company is already capable of reliably manufacturing, into a large-scale quantum processor.
"This modular approach exponentially reduces manufacturing complexity and allows for accelerated, predictable scaling," said the company.
According to Rigetti, this will also enable future systems to scale in multiplicative ways, as individual chips increase their number of qubits, and new technologies enable more of these chips to be connected into larger processors.
With scale being a top priority for virtually every organization in the quantum ecosystem, Rigetti's new launch could well give the startup a competitive advantage, even in an industry crowded with tech giants the likes of Google, IBM, Microsoft and Amazon.
IBM recently unveiled a roadmap for its quantum hardware that aims to build a 1,121-qubit device for release in 2023.
And smaller players are now emerging, often with the goal of exploring alternatives to superconducting qubits that might enable quantum computers to grow faster. UK start-up Quantum Motion, for instance, recently published the result of an experiment with qubits on silicon chips.
"There is a race to get from the tens of qubits that devices have today, to the thousands of qubits that future systems will require to solve real-world problems," said Amir Safavi-Naeini, assistant professor of applied physics at Stanford University. "Rigetti's modular approach demonstrates a very promising way of approaching these scales."
As demonstrated by Rigetti's latest announcement, new approaches, methods and technologies are constantly developing in the quantum ecosystem. It is unlikely that one clear way forward will stand out anytime soon.