The Meta Hired Team is responsible for building Graphcore’s AI networking technology
Meta Platforms has hired an Oslo-based team that until late last year was building artificial intelligence networking technology at UK unicorn chip company Graphcore.
A Meta spokesperson confirmed the hires in response to a request for comment, after Reuters identified 10 people whose LinkedIn profiles said they worked at Graphcore until December 2022 or January 2023 and later joined Meta in February or March of this year.
“We recently welcomed a number of highly specialized engineers to Oslo to our infrastructure team at Meta. They bring deep experience designing and developing supercomputing systems to support AI and machine learning at scale in Meta data centers,” said John Carville. Meta Spokesperson.
The move brings additional strength to the social media giant’s attempt to improve how its data centers handle AI work, as it races to meet demand for AI-oriented infrastructure from teams across the company looking to build new features.
Meta, which owns Facebook and Instagram, has increasingly relied on AI technology to target ads, select posts for its app feeds, and remove banned content from its platforms.
Moreover, it is now rushing to join competitors like Microsoft and Alphabet’s Google in launching generative AI products capable of creating human-like writing, art, and other content, which investors see as the next big growth area for tech companies.
The job descriptions for the 10 LinkedIn employees indicated that the team had worked on AI networking technology at Graphcore, which develops computer chips and systems optimized for AI work.
Carville declined to say what they would be working on on the meta.
Graphcore closed its Oslo office as part of a broader restructuring announced in October last year, a spokesperson for the startup said, as it struggled to make a headway against US-based firms such as Nvidia and Advanced Micro Devices that dominate. on the artificial intelligence chip market.
Two sources told Reuters that Meta already has an in-house unit that designs several types of chips aimed at speeding up and maximizing efficiency for its AI work, including a network chip that performs a kind of air traffic control function for servers.
Efficient networks are especially useful for modern AI systems such as those behind the chatbot ChatGPT or the image builder Dall-E, which are too large to fit into a single computing chip and must instead be divided into many chips strung together.
A new class of network chipset has emerged to help keep data moving smoothly within these computing clusters. Nvidia, AMD, and Intel all manufacture these network chips.
In addition to its network chip, Meta is also designing a complex computing chip to train AI models and perform inference, a process in which trained models make judgments and generate responses to prompts, though it doesn’t expect that chip to be ready until around 2025.
Graphcore, one of the UK’s most valuable start-up technology companies, was seen by investors such as Microsoft and venture capital firm Sequoia as a promising potential competitor to NVIDIA’s flagship AI chip systems market.
However, it faced a setback in 2020 when Microsoft canceled an early deal to buy Graphcore chips for its Azure cloud computing platform, according to a report from British newspaper The Times. Instead, Microsoft used Nvidia’s GPUs to build the massive infrastructure that supports ChatGPT OpenAI developer, which Microsoft also supports.
Sequoia has since written its investment in Graphcore down to zero, though it remains on the company’s board of directors, according to a source familiar with the relationship. The downgrade was first reported by Insider in October.
A Graphcore spokesperson confirmed the setbacks, but said the company is “ideally positioned” to benefit from accelerating commercial adoption of AI.
© Thomson Reuters 2023