DrFacebook’s Meta Group has launched a new supercomputer for artificial intelligence (AI) applications. On Monday, the company announced that the “AI Research Super-Cluster” will be the world’s fastest artificial intelligence computer when it is fully developed in the middle of the year. Meta researchers are already using computers to make digital systems understand what is happening and what is being said in the videos. In the future, the technology could also be used, for example, for simultaneous interpretation in video conferencing or games.
New applications rely on AI models that learn from examples in order to understand languages independently, for example. The models in RSC are designed to learn from trillions of examples and work in hundreds of different languages. This way, text, images, and videos should be able to be analyzed together seamlessly. When it comes to the goal set by company founder and CEO Mark Zuckerberg to align the company with a more inclusive virtual world (“metaverse”), new augmented reality tools must also be created. The computer aims to make a significant contribution to making the metaverse concept a reality for the company.
AI supercomputers are created by combining multiple GPUs into computing nodes, which are linked together via a high-performance network. According to Meta, the RSC currently consists of a total of 760 NVIDIA DGX A100 systems as compute nodes. That’s over 6000 GPUs in total. The storage layer has a capacity of 175 petabytes – that’s 175 million gigabytes. The computer is ready for use, but it will be expanded further. By the middle of the year, the number of GPUs is expected to climb to 16,000, which is said to increase AI training performance by more than 2.5 times.
