We are pleased to announce that the new GPU SuperComputer service has been launched. The service speeds up the processes of deep learning of neural networks 140 times in comparison with calculations on regular processors. Large businesses will have the opportunity to reduce the time spent on development and implementation of projects based on artificial intelligence and big data. Technology start-ups will reduce the cost of high-performance computing related to video processing, face recognition and speech technology by a hundred times.
For example, companies from aerospace and automotive industries, metallurgy industry and the oil and gas sector will significantly save costs and be able to carry out complex engineering calculations and three-dimensional modeling many times faster. In fintech supercomputing can be used to develop solutions for remote identification or credit risk assessment. Retailers will be able to analyze faster the purchasing behavior for more effective sales management.
The GPU SuperComputer service can also be helpful to technology start-ups and small companies due to several pricing options and the possibility of daily use of a supercomputer starting from less than 1 thousand rubles per day. Cost of service for large businesses is subscription based and varies from 29 000 rubles and up to 200 000 rubles monthly depending on the required computing power.
"Nowadays computing on regular servers is gradually giving way to high-performance systems or supercomputers. With the launch of our new service we have opened our customers access to high-performance cloud computing that was previously impossible or required a lot of time and cost. Now businesses will be able to introduce innovations faster thus increasing its competitiveness. We offer a solution, with higher performance, that is three to four times more affordable than Russian counterparts and, compared to foreign ones, five or more times", comments Oleg Molotilov, director of the #CloudMTS cloud provider.
The supercomputer solves problems that involve huge amounts of data due to computing elements in parallel working. The performance of the GPU SuperComputer service reaches one petaflops. This is comparable to performance of more than one hundred regular servers.