Deeptech/Quantum/News/

Quantum algorithms that speed up banking operations 100x are here

Spanish startup Multiverse is supplying BBVA and Bankia with the software 

By Maija Palmer

Multiverse's senior team: Samuel Mugel, CTO; Román Orús, Chief Scientific Officer; and Enrique Lizaso Olmos, CEO.

Forget what Goldman Sachs said about useful quantum computing being five years away in finance. Banks can already get a 100-fold advantage by using quantum computers to solve problems such as portfolio optimisation and fraud detection fraud, says Spanish startup Multiverse Computing.

The company, which raised a €10m seed funding round today, has developed a quantum software product that it is supplying to customers including BBVA, Bankia the European Tax Agency and the Bank of Canada.

“For problems like optimising investment portfolios and detecting fraud, quantum computers can already outperform classical computers.”

“It depends on the problem you are trying to solve,” says Enrique Lizaso, chief executive.  “For some problems like optimising investment portfolios and machine learning to detect banking fraud, quantum computers can already outperform classical computers today.”

Tasks like these can be done 100 times faster using quantum rather than classical computers, even though today’s quantum computers are still relatively limited, with less than 100 qubits and high error rates.

Goldman Sachs — which has been working with quantum startup QCWare to test out practical applications has estimated that it would take machines with 7500 qubits and five more years— for quantum computing to be of practical use to the financial services industry. Jeremy O’Brien, CEO of photonic quantum computing company PsiQuantum, says quantum computers must reach 1m qubits — something he believes is a decade away— to be useful.

However, by focusing on problems that are particularly well suited to quantum computers, Multiverse says it is possible to get quantum advantage even with today’s small, error-prone machines.

The San Sebastian-headquartered company matches the algorithm, on the back end, to a particular type of quantum computer that is best suited for that problem. For example D-Wave’s machines are good for optimisation problems while IBM and IonQ’s quantum computers perform better on machine learning, says Lizaso.

Other types of problems can still be tricky to solve with quantum computing, such as working out pricing for complex financial instruments (which relies on running so-called Monte Carlo simulations that QC Ware and Goldman Sachs were looking at) or for large-scale simulations. For these, we are likely to need quantum machines with larger numbers of qubits.

“Imagine an early computer in the 1970s and asking it to be able to recognise voice and search the internet — it would be impossible. But you could use it to run a spreadsheet.”

“Imagine an early computer in the 1970s and asking it to be able to recognise voice and search the internet — it would be impossible. But you could use it to run a spreadsheet, it would still be useful. That’s what it is like with quantum computers today,” says Lizaso.

Given the fierce competition in the industry to hire top talent and to win relationships with the biggest clients, Multiverse wants to grow quickly. The company is aiming to expand the service into other sectors such as energy and grow headcount from the current 27 staff to close to 50 by the end of this year, and to more than 200 by 2027. The company is planning to raise a further €50m next year.

“You have to grow like hell at the moment.”

“You have to grow like hell at the moment,” says Lizaso, noting that many of Multiverse’s competitors are large tech incumbents like IBM and Google, and that quantum software rival Cambridge Quantum Computing recently merged with Honeywell to give itself more firepower.

Multiverse’s seed round was led by JME Ventures and also included Quantonation, EASO Ventures, Inveready, Mondragón Fondo de Promoción, Ikerlan, LKS, and Penja Strategy.

Join the conversation

avatar
  Subscribe  
Notify of