It is supposed to understand hundreds of languages and be able to process trillions of data: By summer, Meta Platforms, parent company of Facebook, Instagram and WhatsApp, plans to build the world’s most powerful computer for AI applications. The company announced this on Tuesday night.
Meta was silent on the location or cost of the AI Research Supercluster (RSC). Instead, the company announced impressive technical details. The supercomputer will have about 16,000 Nvidia graphics processors, 175 petabytes of storage and a network speed of 200 gigabits per second. The speed in certain computing processes: almost five exaflops, ten times faster than the most powerful supercomputer to date.
With the computer, Facebook wants to advance its own research in the field of artificial intelligence (AI). The goal is to “develop entirely new AI systems that, for example, enable real-time translations for large groups of people who all speak different languages,” according to a blog post from the group. The technology could enable those people to collaborate on a research project or play a game together without communication problems, it said.
The computer is expected to pay dividends on the group’s key project for the future.
Meta raises the bar for innovation
But to enable such experiences with the new computer will take several years, experts say.
They’ll have their work cut out for them, though. For example, he said, Meta is working on AI that pays attention not only to spoken language, but at the same time to body posture, lip movement and other factors – to enable real-time, accurate translations into other languages. “With the computer, Meta is raising the bar for innovation,” Dekate says.
Some applications are much closer: reviewing the posts and comments on Facebook or Instagram. Much like a human, the new supercomputer should be able to build meaningful context from multiple input sources such as language and images.
That would be a great help for the company, which has come under heavy criticism in recent months. For example, the AI repeatedly confused car accidents with cockfights, to the amazement of employees, the Wall Street Journal reported. According to whistleblower documents, the AI detects only three to five percent of inflammatory posts and 0.6 percent of violence and sedition.
But this task is central to Facebook’s business model; the quality of the network and user trust are essential. Then there are the costs. According to internal documents, Facebook spent more than $100 million in 2019 to vet content through employees and surveys.
Artificial intelligence is greatly changing the demands on computers. They have to perform many calculations simultaneously, but less accurately than before – for example, to correctly classify an object in a photo, but not details such as color or the like.
The Group has recognized the importance of hardware
Graphics processors are ideal for this purpose. A graphic processing unit (GPU) is otherwise used for video games, for example, to put as many pixels per second as possible on the screen.
The fact that Meta is developing its own supercomputer probably has another reason as well. The company has learned from the past that it is dangerous to rely only on software development in the competition between the world’s most powerful companies.
Meta is therefore also putting a lot of resources into developing its own glasses for VR (virtual reality) and AR (augmented or augmented reality). The big challenge at the moment is to make the devices smaller, lighter and thus suitable for everyday use.