A Silicon Valley start-up, Cerebras, has revealed its towering ‘supercomputer’ engineered specifically for artificial intelligence research. Standing at 1.8 meters tall, this behemoth runs on the company’s custom chips, which are 56 times larger than the conventional chips used in AI technologies, as reported by The New York Times.
Andrew Feldman, CEO of Cerebras, explained that the company’s goal is to show that supercomputers designed for artificial intelligence can be both faster and more cost-efficient, even when produced by start-ups.
The past year witnessed a surge in artificial intelligence applications, leading to an unparalleled demand for chips.
This demand has prompted technology giants such as Microsoft, Meta, and Google, along with various start-ups and smaller entities, to delve deeper into the AI sector. The AI products these companies develop require significant processing power and specialized chips, which have recently been in short supply.
To combat the supply-demand mismatch, major tech corporations like Google, Amazon, and Intel have developed their own chip alternatives.
Meanwhile, start-ups such as Cerebras, alongside Graphcore, Groq, and SambaNova, have entered the arena primarily dominated by Nvidia, offering their innovative solutions. Yet, the emergence of such computers from start-ups remains an uncommon sight.
Feldman remains optimistic about the AI industry’s future, pointing out that numerous companies are looking to diversify their AI dependencies. They are increasingly considering firms like Cerebras for their chip requirements, moving beyond a sole reliance on Nvidia.
You may also like this content
- Meta Building World’s Fastest AI Supercomputer for Metaverse
- Artificial Intelligence Will Make Decisions Instead Of People
- Bill Gates: Artificial Intelligence Over Web3 and Metaverse
-
-
-
-
-
-
-
-
-
-
