Broadening Our Business Scope to Build the Next Generation of Digital Infrastructure:
Mitsubishi HC Capital’s Pursuit of GPU Sharing and Green Data Center Solutions

―Could you tell us about the structure and role of the Digital Infrastructure Sales Department?

Hamura:
―As the use of generative AI continues to expand in Japan and around the world, demand for the underlying infrastructure, including GPU servers and data centers, is reaching unprecedented levels. At the same time, we’re starting to see issues emerge: semiconductor shortages are driving up procurement costs, and data centers are facing constraints on power supply and physical space. Traditionally, Mitsubishi HC Capital has supported its customers through leasing and financing services. However, we’ve come to realize that, in today's environment, providing financing alone is no longer sufficient to address these issues. We now feel a growing sense of urgency that Japanese companies may be too late to adopt AI on a broad scale.
In April 2026, we established the Digital Infrastructure Sales Department to work toward resolving these issues. The department currently has 12 members, including management, and two sections: the Sales Section and the Business Development Section, where I work. Against the backdrop of growing demand for generative AI and data centers, we are taking on the next frontier of digital infrastructure, which includes GPU servers and data centers.
The Business Development Section is responsible for creating new business opportunities. After several years of preparation, we leveraged the Company's Innovation Investment Fund* to form strategic partnerships with HIGHRESO Co., Ltd. in November 2025 and Morgenrot Inc. in May 2026. Today, we are working with these partner companies to develop new businesses while connecting internal and external expertise and capabilities across organizational boundaries.

―How do your partnerships with the two companies help you address these issues?

Hamura:
To establish a strong presence in this field, it’s essential to work with partners that possess a high level of expertise in cutting-edge technologies. In that regard, our partnerships with both companies are highly significant.

HIGHRESO provides the computing infrastructure required for generative AI through its GPU cloud services and dedicated GPU data centers.
Our collaboration with the company focuses primarily on the development of distributed regional data centers. This concept involves establishing and operating data centers across multiple regional locations, offering benefits such as integration with renewable energy sources and the diversification of disaster-related risks. Data centers capable of supporting large-scale computational workloads have become indispensable for utilizing generative AI and executing advanced simulations. HIGHRESO possesses extensive expertise in the design and operation of data centers that can meet these demands. In particular, the company has developed know-how for effectively repurposing existing facilities in regional areas and designing infrastructure optimized for specific use cases. This approach enables the construction of data center environments that deliver the required performance while accommodating both market demand and facility constraints. As a result, infrastructure can be deployed quickly and at a low cost.

Meanwhile, Morgenrot provides the computing resources (GPUs) required for generative AI and services to manage them. It’s also engaged in developing and operating AI data centers. The company possesses technologies that enable users to monitor the usage status of GPUs and optimize their operations, as well as virtualization technologies for sharing and utilizing unused computing capacity more effectively. It has also obtained international patents in this field. Together with Morgenrot, we are working on the development of GPU visualization and sharing solutions. High-performance GPUs can cost more than ¥100 million per unit, making them highly valuable assets. As a result, maximizing utilization and putting the chips to use in different ways after they’ve been deployed can be significant challenges for users. By owning and providing GPU servers ourselves, Mitsubishi HC Capital can help customers access the computing resources they need for AI development without making substantial upfront investments.

―What synergies can be created through collaboration among the three companies?

Hamura:
By combining HIGHRESO’s expertise in building and operating GPU infrastructure, Morgenrot’s technologies and services for optimizing the utilization of computing resources, and Mitsubishi HC Capital’s asset ownership and financing capabilities, we can provide an integrated solution that delivers both cutting-edge technology and a stable supply framework.
We also believe that these partnerships have the potential to support the expansion of business domains outlined in our medium-term management plan. Looking ahead, in addition to our traditional financing and asset provision services, we want to explore new business opportunities in the field of AI data centers. While specific business models remain under consideration, we believe that establishing an asset utilization model that takes into account an asset’s entire lifecycle could enable us to offer customers more flexible and competitive services. Rather than simply providing physical assets, one of our key objectives is to create a framework in which the computing resources required in the AI era can be efficiently shared, circulated, and utilized. To realize this vision, we will continue exploring new possibilities by leveraging the expertise and strengths of each company.

―How do you approach environmental sustainability in this business?

Hamura:
Among leading companies in Japan, the idea that businesses should take responsibility for the CO₂ emissions generated through their operations is already well established. Looking ahead, the CO₂ emissions associated with the use of digital tools and computing resources are also expected to become an increasingly important management challenge. At the same time, competition is intensifying among digital infrastructure providers for access to power, and securing renewable energy could become a key competitive advantage.
By combining the Mitsubishi HC Capital Group’s capabilities in developing and supplying renewable energy, as well as its expertise in real estate utilization, with the technology and know-how of our partner companies, we believe there’s significant potential to create new value. One example is the concept of linking renewable energy sources with distributed regional data centers to establish environmentally sustainable digital infrastructure. While specific business models are still under consideration, this is one of the future visions we are exploring in the medium to long term.

―How have frontline sales teams responded to these initiatives? Have you seen any changes in motivation within the Company?

Hamura:
In the past, when we introduced AI infrastructure solutions to the finance or general affairs departments of our customer companies, it was not uncommon to hear responses such as, “What exactly is that?” However, over the past few years, as the use of generative AI has expanded, interest has grown significantly, particularly among companies engaged in advanced AI development, as well as manufacturers and research institutions that require advanced simulations and AI applications. Our capital and business alliances with HIGHRESO and Morgenrot have also attracted considerable attention from customers as a clear demonstration of Mitsubishi HC Capital’s commitment to the AI infrastructure sector. The expertise and technological capabilities these companies possess in GPU infrastructure operations and computing resource utilization have earned a high level of trust in the marketplace.
Within the Company, we have seen an unprecedented level of enthusiasm and proactive engagement across departments. As employees follow the rapid pace of developments in AI technologies and applications, many recognize the need to embrace change and explore new possibilities. We regularly receive suggestions and inquiries from frontline sales teams such as, “We have a customer facing this challenge,” or “Could an underused property be transformed into a data center?” These conversations reflect a strong sense of momentum and a willingness throughout the organization to identify new business opportunities.

―Could you share your roadmap going forward and the future you hope to create?

Hamura:
Our first short-term goal is to establish the digital infrastructure business as a solid earnings pillar during the current medium-term management plan period. This market is evolving rapidly, and speed is critical to success. We are determined to acquire the necessary assets and launch full-scale services within FY2026.
Looking further ahead, our vision extends beyond simply providing physical equipment and assets. We aim to deliver new solutions that help improve our customers’ productivity, including physical AI, which integrates AI with real-world assets such as factories and robots, and local large language models, which enable companies to use generative AI securely within dedicated private environments.
To make this vision a reality, we will continue collaborating with a wide range of partners, including startups, to complement areas where additional technological expertise and capabilities are required and to create new value together. By combining our financial capabilities and asset management expertise with the technologies and services of our partner companies, we aspire to help advance AI adoption and build the digital infrastructure that supports a sustainable society in Japan.