Software Specialist Horizon Quantum to Build First-of-a-Kind Hardware Testbed – HPCwire

Horizon Quantum Computing, a Singapore-based quantum software start-up, announced today it would build its own testbed of quantum computers, starting with use of Rigettis Novera 9-qubit QPU. The approach by a quantum software specialist to build-its-own testbed is new. The idea is to be able to develop and integrate its software stack Triple Alpha more thoroughly into various types of quantum computers.

Founded in 2018, Horizons broad strategy is to develop tools that will take software developed using current programming languages and translate that code into quantum algorithms and specific device codes across multiple quantum qubit modalities. The Novera QPU is a superconducting qubit, but Horizons plans call for integrating other qubit modalities into its testbed.

In interview with HPCwire, Horizon CEO and founder, Joe Fitzsimons, said We have been pursuing an ambitious plan to bridge the gap between conventional software engineering and quantum computing through the automation of quantum algorithm construction. Our goal is to enable software engineers and domain experts in fields that make significant use of high performance computing to develop code using familiar programming languages and automatically accelerate these programs using quantum processing. We have already been able to demonstrate automated construction of quantum algorithms from programs written in a subset of the Matlab language, and we expect to integrate such functionality into our development tools over time.

The testbed, which will be set up at Horizons Singapore headquarters, will have the capacity to host multiple quantum computers. In addition to using Rigettis Novera 9-qubit QPU, Horizon will also use Quantum Machiness OPX1000, the processor-based quantum controller. This first integrated system is expected to be installed by early 2025.

Horizon reported in the official release, By acquiring its own hardware, Horizon gains full control over both hardware and software stacks, allowing it to push the frontiers of quantum computing. A key aspect of Horizons quantum computing testbed is its modular multi-vendor approach. Rather than utilizing a single-vendor solution, the company has purposely selected best-in-class components from different providers. This modularity allows Horizon to integrate its software stack with different hardware configurations and upgrade the system over time.

Asked why isnt everyone doing this?

Fitzsimons said, The answer is partly that the timing hasnt been right until now. As we get closer to seeing practical error-corrected quantum computation, the timeline to useful quantum computation is accelerating. While we may well be the first quantum software company to make such a move, I doubt very much that we will be the last.

Its interesting to note the international flavor of the supply chain here. Rigetti, of course, is a U.S.-based quantum computing pioneer. Quantum Machines, founded in 2018, is an Israel-based startup specializing in quantum control systems. Horizon is one of many young and ambitious Asia-PAC based quantum companies. It completed series A funding round ($18 million) roughly a year ago. The global nature of the quantum computing supply chain has basically become a reality.

Like most quantum start-up CEOs, Fitzsimons background is in the science. His Ph.D. (Oxford) is in quantum computing architectures. In 2018 he held a tenured position as an associate professor at the Singapore University of Technology and Design, where he led the Quantum Information and Theory group. He was also a principal investigator at the Centre for Quantum Technologies (CQT), which was established in December 2007 by Singapores National Research Foundation and Ministry of Education, and is hosted by the National University of Singapore.

Fitzsimons told HPCwire, We will be building the system from components ourselves, and expect to have the system operational in early 2025. We will be integrating the system with our software development tools, which enable far more complex programs than many existing quantum programming frameworks since they enable non-trivial flow control and concurrent classical and quantum computation. We expect to open the system up to users of our tools once the integration is complete.

He declined to say which modalities will be brought into its testbed next, We have been very conscious of the significant progress across a number of modalities in the past twelve months. As we get closer to useful quantum computation, we want to ensure that we build up the experience of integrating with, and potentially operating, quantum computers based on the most promising modalities. We will be closely monitoring progress across the field, but will only be making a decision on further systems after the first quantum computer is operational.

On the whole, the Horizon gambit is interesting. It will be interesting to watch the extent to which future systems are brought as components or complete systems. Quantum Machines, on its website, lists several modalities that its control systems can work with, including superconducting, optically addressable (e.g. NV diamonds), quantum dots, and neutral atoms. The move is also interesting for Rigetti, which just entered the merchant QPU market back in December the Novera kit list price then was $900,000.

Included in the official announcement were quotes from Rigetti and Quantum Machines:

Asked about collaborations and working with other AsiaPAC companies, Fitzsimons said, Our main focus is on working with hardware partners, and to date these have been based in North America and Europe. The focus is on pushing forward towards useful quantum computing, and working with other companies that share that goal. We have access to quite a number of systems both through the major cloud providers and through direct access with hardware companies, and have integrated many of these into our tool chain so that users can not only develop quantum programs, but also deploy these programs as APIs which execute jobs on both hardware and simulator backends.

Fitzsimons seems a realist in terms of challenges ahead and uncertainty around the timeline to deliver quantum advantage.

The biggest challenges for any quantum computing company are correctly pacing resource utilization pre-quantum advantage and the limited pool of scientists with significant experience in the field, he said.

One the timing to quantum payoff, he added, I have never been a big believer in the likelihood of really useful quantum computing emerging from variational algorithms used on NISQ machines. Over the past 18 months, however, there has been tremendous progress in error correction and fault-tolerance, and we are seeing an increasing number of experiments exceed breakeven error correction. Over the next three years, I would expect to see significant progress towards the low noise regime.

Stay tuned.

Link to announcement, https://www.hpcwire.com/off-the-wire/horizon-quantum-computing-to-pioneer-multi-vendor-quantum-hardware-testbed/

Excerpt from:

Software Specialist Horizon Quantum to Build First-of-a-Kind Hardware Testbed - HPCwire