How Qumulo helped Japan’s Hirosaki University forge ahead in AI research and education
One of the country’s pioneering institutions in AI and medical imaging, the institution modernized its file data platform with Qumulo to teach today’s students skills for tomorrow.
Nestled on the northern tip of Japan’s main island, Hirosaki University lies in the shadow of snow-covered Mount Iwaki in the medieval castle town of Hirosaki City, host of one of the country’s most renowned cherry blossom festivals.
Although the institution’s setting is ancient, Hirosaki University’s Faculty of Science and Technology is conducting truly cutting-edge research. While Artificial Intelligence (AI) has recently become a buzzword, researchers at Hirosaki University have been developing practical uses for this technology for nearly two decades. Their work includes innovative projects such as translating the dialect of the Tsugaru region into standard Japanese and analyzing data from the Shirakami Mountains, a UNESCO World Heritage site, using satellite images and aerial photographs.
While AI has been a critical component of Hirosaki University’s academic research for years, it has become necessary to educate students so that they can learn and effectively apply AI technologies in their educational curriculum.
To address the need for accessing and managing the massive datasets required for AI training while simultaneously cultivating AI expertise among its incoming students, Hirosaki University needed an infrastructure capable of handling large amounts of image files to support high-capacity data analysis.
One Platform for AI Research and Education
Working with NTT EAST Inc. and Arch Systems Inc. as the integrator and system architect for overall design and implementation, Hirosaki University selected Qumulo to serve both researchers and students.
Designed to address the unique challenges of modern data storage, including AI and machine learning data processing, Qumulo accelerates data workflows, boosts productivity, and improves responsiveness. In addition, it reduces costs by simplifying and enhancing storage and data management.
The university deployed one Qumulo cluster to host data used by faculty and students at the Faculty of Science and Technology for educational and research purposes. A second cluster was deployed at the institution’s Information Technology Center to provide computing infrastructure, services, and support to the school’s five undergraduate faculties and nine graduate schools of the university.
Each cluster effectively handles high-performance computing workloads as well as data archiving and ingestion, simplifying data management while providing optimal storage capacity and performance to benefit multiple educational disciplines.
After hardware installation, integrators from the IT department conducted detailed discussions with Qumulo engineers to ensure the success of the initial implementation and to plan and implement the process of migrating data from the university’s legacy systems, while maintaining the existing permissions structure for both Windows and Linux users.
Now, as Hirosaki University forges ahead in both AI-enabled research and educating incoming students about the burgeoning field of AI algorithms and large language models, it has the storage and compute infrastructure necessary to discover new uses for AI.
A Modern Approach to Customer Success
Previously, when any questions or issues would arise concerning their legacy storage system’s configuration or performance, the university’s IT team had to send an email to their vendor’s support team and then await a reply, which often took several hours or even days.
One of the key differentiators that led Hirosaki University to select Qumulo is Qumulo Zero Latency support, a unique customer success-oriented program that allows customers to engage directly with Qumulo support engineers via Slack, Microsoft Teams, email, or phone, and receive instantaneous responses to resolve questions and in-flight problems in real-time.
Once Hirosaki University’s Qumulo clusters were operational, their IT team was able to engage directly with Qumulo support engineers via Slack, significantly reducing the time to answer questions and resolve issues.
Now, as Hirosaki University forges ahead in both AI-enabled research and educating incoming students about the burgeoning field of AI algorithms and large language models, it has the storage and compute infrastructure necessary to discover new uses for AI, while producing Japan’s next generation of AI leaders for tomorrow.
Business Impact
- Delivered a modern, distributed unstructured data platform that could meet Hirosaki University’s capacity, performance, and data management demands associated with large language models, machine learning, and AI algorithms.
- Enabled Hirosaki University to expand its curriculum to offer students ongoing AI-focused research on the development and use of AI applications.
- Streamlined access from Windows and Linux systems, while protecting security permissions, enabling legacy and modern containerized applications to access and share data for research purposes.
- Allowed direct communication with Qumulo engineers through Slack to ensure initial deployment success and ongoing operations, eliminating the time lag associated with their existing vendor’s support.
Industry:
Higher Education & Research
Use Case:
AI and data science research, high-performance computing, multi-platform access
Deployment:
Two Qumulo clusters supporting research and education workloads
Company Overview:
Hirosaki University is a national research institution located in northern Japan, recognized for its pioneering work in artificial intelligence, medical imaging, and interdisciplinary science. The university plays a leading role in developing AI-based applications such as disease prediction models, language translation, and satellite data analysis.
Why Qumulo:
- Needed a scalable, high-performance platform to manage growing datasets used in AI research and education
- Required seamless access from both Windows and Linux environments while maintaining data security
- Sought real-time, responsive support to streamline operations and reduce IT overhead
Benefits:
- Supports both faculty research and student coursework in AI and data science
- Seamlessly integrated with the university’s legacy systems while modernizing performance
- Enables direct communication with Qumulo engineers via Slack for faster issue resolution
- Empowers the next generation of AI talent through reliable, scalable data infrastructure