A scientist at Harvard University has taken an innovative approach to address the shortage of powerful computing resources for a heart disease study by utilizing Google’s cloud platform to clone a supercomputer. This novel move could potentially pave the way for other researchers to overcome computing limitations and accelerate their work.
The study conducted by Harvard aimed to simulate a therapeutic procedure involving the dissolution of blood clots and tumor cells in the human circulatory system. This simulation required an extensive amount of computing power typically accessible through a supercomputer. However, limitations in supercomputer availability posed a challenge.
Professor Petros Koumoutsakos of Harvard explained that the initial simulation could be executed using a full-scale supercomputer, but refining and optimizing the simulation demanded further access to the same powerful computing resource.
Due to the scarcity of supercomputers capable of handling the complex calculations needed for such research, bottlenecks emerged in the scientific process. In response, researchers and companies, including Citadel Securities, have turned to public cloud services to access the substantial computing resources required.
Although cloud computing operations are not originally designed to meet the specific demands of complex scientific computations, they offer the potential to overcome limitations and enhance productivity. Google Cloud’s Chief High Performance Computing Technologist, Bill Magro, highlighted the transformation of cloud infrastructure to emulate a supercomputer, necessitating adjustments in software, networking, and hardware design.
This innovative approach showcases the adaptability of cloud platforms in addressing scientific and engineering challenges, unlocking productivity and insights in a faster and more efficient manner.
