Meta Begins Testing its First in-house AI Training Chip

The Meta logo, a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
The Meta logo, a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
TT
20

Meta Begins Testing its First in-house AI Training Chip

The Meta logo, a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
The Meta logo, a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo

Facebook owner Meta (META.O), opens new tab is testing its first in-house chip for training artificial intelligence systems, a key milestone as it moves to design more of its own custom silicon and reduce reliance on external suppliers like Nvidia (NVDA.O), opens new tab, two sources told Reuters.

The world's biggest social media company has begun a small deployment of the chip and plans to ramp up production for wide-scale use if the test goes well, the sources said.

The push to develop in-house chips is part of a long-term plan at Meta to bring down its mammoth infrastructure costs as the company places expensive bets on AI tools to drive growth.

Meta, which also owns Instagram and WhatsApp, has forecast total 2025 expenses of $114 billion to $119 billion, including up to $65 billion in capital expenditure largely driven by spending on AI infrastructure.

One of the sources said Meta's new training chip is a dedicated accelerator, meaning it is designed to handle only AI-specific tasks. This can make it more power-efficient than the integrated graphics processing units (GPUs) generally used for AI workloads.

Meta is working with Taiwan-based chip manufacturer TSMC (2330.TW), opens new tab to produce the chip, this person said.

The test deployment began after Meta finished its first "tape-out" of the chip, a significant marker of success in silicon development work that involves sending an initial design through a chip factory, the other source said.

A typical tape-out costs tens of millions of dollars and takes roughly three to six months to complete, with no guarantee the test will succeed. A failure would require Meta to diagnose the problem and repeat the tape-out step.

The chip is the latest in the company's Meta Training and Inference Accelerator (MTIA) series. The program has had a wobbly start for years and at one point scrapped a chip at a similar phase of development.

However, Meta last year started using an MTIA chip to perform inference, or the process involved in running an AI system as users interact with it, for the recommendation systems that determine which content shows up on Facebook and Instagram news feeds.

Meta executives have said they want to start using their own chips by 2026 for training, or the compute-intensive process of feeding the AI system reams of data to "teach" it how to perform.

As with the inference chip, the goal for the training chip is to start with recommendation systems and later use it for generative AI products like chatbot Meta AI, the executives said.

"We're working on how would we do training for recommender systems and then eventually how do we think about training and inference for gen AI," Meta's Chief Product Officer Chris Cox said at the Morgan Stanley technology, media and telecom conference last week.

Cox described Meta's chip development efforts as "kind of a walk, crawl, run situation" so far, but said executives considered the first-generation inference chip for recommendations to be a "big success."

Meta previously pulled the plug on an in-house custom inference chip after it flopped in a small-scale test deployment similar to the one it is doing now for the training chip, instead reversing course and placing orders for billions of dollars worth of Nvidia GPUs in 2022.

The social media company has remained one of Nvidia's biggest customers since then, amassing an arsenal of GPUs to train its models, including for recommendations and ads systems and its Llama foundation model series. The units also perform inference for the more than 3 billion people who use its apps each day.

The value of those GPUs has been thrown into question this year as AI researchers increasingly express doubts about how much more progress can be made by continuing to "scale up" large language models by adding ever more data and computing power.

Those doubts were reinforced with the late-January launch of new low-cost models from Chinese startup DeepSeek, which optimize computational efficiency by relying more heavily on inference than most incumbent models.

In a DeepSeek-induced global rout in AI stocks, Nvidia shares lost as much as a fifth of their value at one point. They subsequently regained most of that ground, with investors wagering the company's chips will remain the industry standard for training and inference, although they have dropped again on broader trade concerns.



Trump Offers Support to Musk's Car Company in a Surprising Post as Tesla Stock Plunges

Mar 22, 2025; Philadelphia, PA, USA; Elon Musk and President Donald Trump during the Division I Men's Wrestling Championship held at Wells Fargo Center. Mandatory Credit: Eric Hartline-Imagn Images/File Photo
Mar 22, 2025; Philadelphia, PA, USA; Elon Musk and President Donald Trump during the Division I Men's Wrestling Championship held at Wells Fargo Center. Mandatory Credit: Eric Hartline-Imagn Images/File Photo
TT
20

Trump Offers Support to Musk's Car Company in a Surprising Post as Tesla Stock Plunges

Mar 22, 2025; Philadelphia, PA, USA; Elon Musk and President Donald Trump during the Division I Men's Wrestling Championship held at Wells Fargo Center. Mandatory Credit: Eric Hartline-Imagn Images/File Photo
Mar 22, 2025; Philadelphia, PA, USA; Elon Musk and President Donald Trump during the Division I Men's Wrestling Championship held at Wells Fargo Center. Mandatory Credit: Eric Hartline-Imagn Images/File Photo

President Donald Trump took to social media Thursday morning to support Elon Musk's car company, a startling development given their bitter public feud.

”I want Elon, and all businesses within our Country, to THRIVE,” Trump wrote on Truth Social, Reuters reported.

The post wasn't enough to help Tesla's stock, which fell sharply after the company reported another quarter of lackluster financial results and Musk warned of some potentially “rough quarters” into next year. At midday, the stock was down around 9%.

Late Wednesday, Tesla said revenue fell 12% and profit dropped 16% in the April-June quarter. Many prospective buyers have been turned off by Musk’s foray into right-wing politics, and the competition has ramped up in key markets such as Europe and China.

Investors have been unnerved by Musk's social media spat with the president because Trump has threatened to retaliate by ending government contracts and breaks for Musk's various businesses, including Tesla.

But Trump struck a starkly different tone Thursday morning.

“Everyone is stating that I will destroy Elon’s companies by taking away some, if not all, of the large scale subsidies he receives from the US Government. This is not so!" Trump wrote. “The better they do, the better the USA does, and that’s good for all of us.”

After Trump's massive budget bill passed earlier this month, Tesla faces the loss of the $7,500 EV tax credit and stands to make much less money from selling regulatory credits to other automakers. Trump’s tariffs on countries including China and Mexico will also cost Tesla hundreds of millions of dollars, the company said on its earnings call.

Musk has blasted the budget bill on his own social media platform X for adding to US debt at a time when it is already too large. The Tesla CEO has called the budget pushed by the president a “disgusting abomination” and has threatened to form a new political party.

On Wednesday's call, Musk said the electric vehicle maker will face “a few rough quarters” as it moves into a future focused less on selling cars and more on offering people rides in self-driving cars. He also talked up the company's business making humanoid robotics. But he acknowledged those businesses are a ways off from contributing to Tesla’s bottom line.

Tesla began a rollout in June of its paid robotaxi service in Austin, Texas, and hopes to introduce the driverless cabs in several other cities soon. Musk told analysts that the service will be available to probably “half of the population of the US by the end of the year — that’s at least our goal, subject to regulatory approvals.”

“We’re in this weird transition period where we’ll lose a lot of incentives in the US,” Musk said, adding that Tesla “probably could have a few rough quarters” ahead. He added, though, “Once you get to autonomy at scale in the second half of next year, certainly by the end of next year, I would be surprised if Tesla’s economics are not very compelling.”