Meta Recruit AI Network Development Team from Graphcore

iwanrj.com – Jakarta – Currently, Meta is recruiting a team of developers based in Oslo Norway who at the end of last year built an AI-powered network on Graphcore.

For those who don’t know, Graphcore is a British semiconductor company that develops accelerators for use by artificial intelligence and machine learning technologies.

Jon Carvill Spokesperson for Meta confirmed that the company’s decision to hire the AI ​​development team was in response to the apparent 10 people on LinkedIn who said they worked at Graphcore until December 2022 or 2023 or January 2023 and then joined Meta in February or March this year.

READ ALSO:

“We recently welcomed a number of highly specialized engineers in Oslo to our infrastructure team at Meta. They bring deep expertise in the design and development of supercomputing systems to support AI and machine learning technologies at scale in Meta data centers,” said Carvill.

Reported Telset from Gadgets360this move could provide additional strength to the social media development company’s bid to improve the way data centers handle AI work, as teams race to address the demand for AI-oriented infrastructure to build new features.

Meta, as the owner of Facebook and Instagram, is increasingly relying on AI technology to be able to target ads, select posts for application feeds, to clean prohibited content from the platform he made.

In addition, the company is also in the process of developing generative AI products capable of creating human-like writing, art and other content, following its competitors Microsoft and Google. Generative AI products are seen by investors as the next big growth area for technology companies.

Job descriptions on LinkedIn of the 10 employees indicate that the team has worked on specialized AI network technologies at Graphcore, which develops computer chips and systems optimized for AI work. However, Carvill declined to say what they will be working on on Meta.

On the other hand, Graphcore has closed its Oslo office as part of a broader restructuring announced in October last year, said a spokesperson for the startup, as it struggles to make inroads to compete with companies like Nvidia and AMD that dominate the market. chips AI.

Meta currently has an internal unit that designs several types of chips aimed at accelerating and maximizing the efficiency of AI work, including network chips that perform a kind of server function for air traffic control, according to Reuters.

Efficient networking is particularly useful for modern AI systems such as the one used by ChatGPT, which are too large to fit onto a single computing chip and must instead be split across multiple chip sets.

A new category of network chips has emerged to help keep data flowing smoothly in computing clusters. Nvidia, AMD, and Intel have made such network chips.

In addition to networking chips, Meta also designs complex computing chips to train AI models and perform inference, a process by which a trained model makes judgments and generates responses to commands, although these chips are not expected until around 2025.

In addition to its network chips, Meta also designs complex computational chips to train AI models and perform inference, a process by which a trained model makes judgments and generates responses to commands, though it isn’t expected to have the chip ready until around 2025.

Meanwhile, Graphcore, the UK’s most valuable technology startup, has been eyed by investors such as Microsoft and venture capital firm Sequoia and is seen as a potential challenger to Nvidia in the AI ​​chip systems market.

However, its development suffered a setback in 2020 when Microsoft canceled its initial buyout deal chips Graphcore for cloud computing owned by the company.

After that, Microsoft instead used Nvidia GPUs to build a massive infrastructure that supports ChatGPT developers.

READ ALSO:

On the other hand, Sequoia has since written down its investment in Graphcore to zero, although it remains on the company’s board. The drop in investment was first reported by Insider back in October.

A Graphcore spokesperson confirmed the setback, but said the company is perfectly positioned to take advantage of the accelerating commercial use of AI technology. [NM/HBS]