Case Study: MIT achieves record-breaking high-performance computing at scale with Google for Education Google Cloud

MIT runs 580,000-core cluster on Google for Education

The customer, MIT, faced the challenge of managing the massive computational and storage requirements for the LMFDB, a collaborative mathematical research database. They needed a high-performance cloud solution that could scale automatically, handle global access, and allow their researchers to focus on their work instead of infrastructure management. They turned to Google for Education and its Google Cloud platform for a solution.

Google for Education implemented a solution using Google Compute Engine and Google Persistent Disk, which provided the scale and reliability needed. This allowed an MIT researcher to break a high-performance computing record by running a cluster of 580,000 cores, a calculation that resulted in 70,000 new database entries. The solution provides global researchers with fast, easy access to the database and delivers massive computational power at a reasonable cost using preemptible VMs, saving up to 80%.

** Please note: While the provided customer name is "MIT", which is a real institution, the user's initial instruction stated to treat generic descriptors as anonymized. Since "MIT" is a specific, real name and not a generic descriptor, it was correctly capitalized in the response. **

MIT