Japan's JHPC-quantum selects 20 more projects to test its hybrid quantum and supercomputer platform

[Shane95]

RIKEN and SoftBank have picked 20 new proposals for the Test User Program run by JHPC-quantum, HPCwire reports. JHPC-quantum is a Japanese project building software that links quantum computers with supercomputers into a hybrid platform. This was the programme's third open call, and it ran from 10 May to 30 June. The previous cohort had 21 proposals

Applicants came from industry and academia in Japan and abroad. Fields include materials, chemicals, manufacturing, ICT and healthcare. Proposals were judged mainly on how specifically they would use the platform, how central supercomputing was to the work and how soon results would come. In other words, the organisers wanted real hybrid projects, not quantum experiments that happen to sit next to a supercomputer

The list of lead proposers is impressive. Companies include Mitsubishi Electric, Panasonic System Networks R&D Lab, IHI, Sumitomo Mitsui Banking Corporation, DIC, TISI and Algorithmiq. Research bodies include JAXA, the University of Osaka, Tohoku University, the University of Tokyo and Germany's DLR, as well as RIKEN's own Center for Computational Science. It is a broad mix of heavy industry, finance and space

The hardware on offer is serious. Selected teams get access to the Reimei quantum computer, IBM's Quantum System Two in Kobe, the Fugaku supercomputer and the ROQUO GPU supercomputer. Teams will present at the JHPC-quantum Symposium on 9 November. The final open call is planned for early in Japan's 2027 financial year

Japan's approach here is very practical. Rather than chasing headlines, it is putting real machines in front of real companies and asking them to try things. Fugaku sitting alongside quantum hardware gives the hybrid idea a proper test. I would love to see some of the results shared publicly after the symposium

Is hybrid quantum and supercomputing the most realistic path to useful results? And which of these organisations would you expect to get something useful first?

Japans Jhpc Quantum Expands Test User Program With 20 Selected Proposalswww.hpcwire.com

[Freya_27]

Banks are always first in these programmes. SMBC will probably look at portfolio optimisation or risk. Those are well studied problems with clear classical benchmarks. That makes it easier to see whether quantum is helping. They also have the budgets to experiment

[Hollow Coder]

JAXA being involved is fun. Space agencies have endless optimisation and simulation problems, from trajectories to materials. Even small gains would be worth having

I wonder what their proposal covers. Satellite scheduling is a classic optimisation headache

[HollowSentinel]

Access to Fugaku alone is a big draw. It is still one of the most capable supercomputers around, especially for scientific workloads

Pairing it with IBM's machine in Kobe gives teams a lot to play with. Twenty projects is a decent number. Kobe has become a real hub for this work