Microsoft has created a cloud-based supercomputer, which is one of the five most powerful supercomputers in the world and is exclusively for OpenAI, founded by Elon Musk and former venture capital fund manager Alt Combin Y Combinator. A new supercomputer designed for artificial intelligence (AI) developments were unveiled at the Microsoft Build 2020 conference, which is taking place online this year due to the coronavirus pandemic.
According to a Microsoft blog, the supercomputer combines more than 285,000 processor cores, 10,000 graphics processors and has a network connection at a speed of 400 Gbit / s for each server. By hosting the Azure Cloud service, the supercomputer benefits from the advantages of cloud infrastructure, including rapid deployment, data center reliability, and access to Azure services.
Microsoft noted that before, machine learning experts had developed AI models using data sets to solve a single task: object recognition, text translation, etc. However, a large model is better able to handle certain scenarios, formed, for example, on billions of pages of text accessible to the public. Systems of this type can solve several different tasks at once: summarize the results of long performances, moderate chat in online games, and even generate program code. But their work requires a powerful infrastructure, which is why Microsoft has created a new supercomputer.
“The supercomputer was created specifically to train the next generation of massive AI models capable of performing a wider range of functions and, therefore, requiring large computing resources,” said Microsoft (quote from TASS).
OpenAI is committed to regulating artificial intelligence. OpenAI is also developing tools for working with neural networks and machine learning – for example, an algorithm that generates compelling fake news.