Six of Japan’s big industrial names, SoftBank Corp., TOPPAN Holdings, AISIN, Asahi Kasei, AGC and DIC, have signed a joint agreement to build a platform for finding new materials that combines quantum computing with AI. They announced it on 28 October in a joint release published by SoftBank and TOPPAN, among others.
The work runs from fiscal 2026, which ends on 31 March 2027, to fiscal 2028. It is backed by Japan’s New Energy and Industrial Technology Development Organization (NEDO), the state research funding agency, which chose the project in a funding round this summer. None of the companies gave a budget.
In brief
- Six Japanese companies have signed a joint agreement to build a materials discovery platform using quantum computing and AI.
- The NEDO-backed work runs to the end of fiscal 2028, in March 2029, and targets battery, optical and superconducting materials, among others.
- The companies say quantum computers alone cannot yet handle large calculations, so the plan pairs them with AI, supercomputers and lab tests.
What the six companies want to build
The release says the project “aims to establish a next-generation materials development platform by integrating quantum-computing-based calculations with AI technologies such as machine learning”. Its goal is to bridge “the gap between demonstration and real-world application”.
The companies explain the problem in plain terms. AI tools that predict new materials need large amounts of good training data, which is often missing. Physics-based computer simulation gets very expensive as calculations become more accurate and larger. Quantum computers “have the potential to simulate the properties of molecules and electrons” more efficiently, the release says, but today’s machines have too few qubits and too many errors, so “large-scale calculations using quantum computers alone are still not a reality”.
The plan is to use quantum calculations and AI together, so each covers the other’s weak spots. Rather than chase one material, the six will test the approach across many, and will work with “specialized quantum computing institutions”, which the release does not name.
The work has three parts. First, the companies will check where quantum computing helps, for materials such as those needed for next-generation batteries, optical materials for new displays and sensors, and superconducting materials for future devices. Second, they will develop the combined quantum and AI methods. Third, they will make real materials based on the simulations and compare lab results with predictions to improve accuracy.
Who does what
Each company takes the materials it knows best:
- TOPPAN will run hybrid quantum and classical algorithms on quantum computers for additive materials.
- AISIN, the car parts maker, will study superconducting materials and compare the results with its usual methods. It will also “forecast practical implementation timelines”.
- Asahi Kasei will study electrolytes, polymers and electrode materials.
- SoftBank will use a computing platform that links quantum computers with supercomputers to produce data for materials work, and will run tests on energy materials such as next-generation batteries.
- AGC will work on optical glass and electronic materials and aims to set design guidelines.
- DIC will try to predict the key properties of colourants and semiconductor materials before they are made.
Part of a wider NEDO push
NEDO listed the project and all six companies in July, among nine quantum computing projects chosen for “large-scale demonstrations to create use cases” under a post-5G research programme, according to its selection page. The Japanese versions of the release describe the work as a NEDO subsidy project. NEDO says the programme aims to industrialise quantum computers in Japan around 2030. Other projects on the same list include drug-discovery work by Chugai Pharmaceutical and a battery-materials simulation project by the start-up Quemix.
What is new this week is the signed agreement between the six and the detailed work plan.
Why the six are hedging on quantum hardware
The release is unusually frank about the limits of today’s quantum machines, and the project is built around that. It does not bet on a quantum computer solving a materials problem on its own. It pairs the machines with AI and supercomputers, and checks every prediction in a real lab. That is a sensible design. The same caution runs through Quantinuum and GE Vernova’s new plan for studying superconductors on Quantinuum’s Helios computer, where GE Vernova talks only of learning where quantum machines can help.
The evidence so far is thin. Small chemistry runs, such as a 42-qubit simulation of one molecule’s NMR spectrum, came with the authors’ own warning that classical methods already cover most practical cases. Error rates are falling, as Google’s magic state cultivation result on its Willow chip shows, but large error-corrected machines are still years away. Meanwhile AI alone is moving fast: an AI-guided robotic lab recently reported hydrogen catalysts it called comparable to platinum, with no quantum hardware at all.
That makes AISIN’s task the most telling. A company forecasting when quantum methods will be practical for its own materials work is, in effect, a buyer testing the sales pitch. If the six can show a single case where quantum calculations beat classical ones on a real material, and the lab confirms it, that would matter far beyond Japan. If not, the shared data and AI tools could still be useful, and the honest answer would be that quantum’s turn has not come yet.
What to watch next
The release gives no milestones before fiscal 2028 ends in March 2029. The first signs will be published results, the names of the quantum computing partners and which machines they use. Researchers are also still learning what these machines can do, as Pasqal’s recent 256-qubit magnet experiment shows.
Sources
- SoftBank Corp., “Six Companies Launch Joint R&D to Build a Quantum-Computing-Based Materials Development Platform”, 28 October 2026
- TOPPAN Holdings, “Six Companies Launch Joint R&D to Build a Quantum-Computing-Based Materials Development Platform”, 28 October 2026
- AISIN, “Six Companies Launch Joint R&D to Build a Quantum-Computing-Based Materials Development Platform”, 28 October 2026
- AGC, joint release (in Japanese), 28 October 2026
- NEDO, selection of implementers for the post-5G quantum computer programme (in Japanese), July 2026
- NEDO, Annex 2: list of selected projects (in Japanese, PDF)
Related stories
- Google Quantum AI Shows “Magic State Cultivation” Working on Its Willow Chip, Cutting Errors About 40-Fold
- Quantinuum Says GE Vernova Will Use Helios to Study Superconductors
- Quantum Roundup: A 42-Qubit Molecule Simulation, a Chip That Holds Photons, and Oratomic’s $775 Million
- AI-Guided Lab Finds Hydrogen Catalysts It Calls Comparable to Platinum
Follow @tsnmediahub on X for breaking AI news as it happens.
Follow on X
