One of the world’s largest supercomputers survived for only 10 minutes. The virtual supercomputer, developed by YellowDog, a cloud workload management (CWM) specialist from the United Kingdom, had an extremely short lifespan. The supercomputer had clocked 3.2 million virtual CPUs (vCPUs) by the time it was dying. The dead supercomputer is therefore among the top ten fastest supercomputers in the world – comparable to Fugaku, but only for a few minutes.
Costs were about $65,000
A well-known drug discovery application was deployed by a pharmaceutical company on a cluster for a single project. The project cost about $65,000. In the past, supercomputers were a privilege and remained in the hands of only a few private organizations, governments, and scientific groups. Cloud computing and CWM tools have made supercomputers more accessible. As TechRadar reported, the supercomputer ran 33,333 instances of AWS’s 96-core c5.24xlarge instance at $1.6013 per hour, comparable to dedicated servers. The session lasted 65 minutes.
Read more: Who plans cities in India? Developers who adhere to colonial masterplans
In just 7 hours, AWS’s ‘on-demand supercomputer’ analyzed 337 million compounds, according to Colin Bridger. A localized system would have taken two months to complete the same process. Among the hardware needed for these supercomputers are web hosting, cloud storage, and website building. These supercomputers are now accessible to anyone who can shell out the cash. With substantial developments in CWM platforms, complex functions can now be performed utilizing machine learning and algorithms.
Post Your Comments