In an unprecedented move, a Harvard scientist has leveraged Google’s cloud platform to clone a supercomputer, thereby significantly accelerating his research on heart disease. This creative solution can potentially guide other researchers in their quest to circumvent the scarcity of robust computing resources.
Professor Petros Koumoutsakos was at the forefront of a study focusing on a therapy designed to break down blood clots and tumor cells in the human circulatory system. However, the enormous computing power required to run simulations of this nature can generally only be accessed through supercomputers.
The challenge was that “we could run one simulation using a full-scale supercomputer,” Koumoutsakos explained, but optimizing and refining the simulation demanded additional access to the supercomputer. Furthermore, in the U.S., the limited number of supercomputers capable of executing the complex calculations required for Koumoutsakos’s study created significant bottlenecks in the research process.
Recognizing this problem, some researchers and corporations like Citadel, which require extensive computing resources, began exploring public cloud options. However, typical cloud computing operations are not crafted to handle the substantial demands researchers have.
Despite being built for individual, smaller computing tasks like video streaming or web page serving, cloud technology has been catching the eye of the scientific community. “Folks are realizing the potential for cloud to solve problems and technical scientific engineering computing to really unlock productivity and get to better answers, better insights, faster,” said Bill Magro, chief high-performance computing technologist at Google Cloud.
Nevertheless, adapting cloud infrastructure to mimic a supercomputer is no small feat. It entails modifications in software, networking, and the physical hardware design. Citadel, in collaboration with Alphabet Inc’s subsidiary Google, sponsored Koumoutsakos’s trailblazing research.
In summary, the novel use of Google’s cloud platform to emulate a supercomputer by a Harvard scientist illustrates a promising avenue for future scientific exploration. This technology’s adaptability may provide a pathway for researchers worldwide to expedite their work, opening doors to new discoveries and innovations.
This news is based on the article from thestar.com.my.
