Meta has unveiled a roadmap for four new in-house chips as part of its MTIA program, with the first, the MTIA 300, already powering its ranking systems. The company is focusing heavily on inference chips to handle exploding demand from AI models responding to user queries, though it has struggled to develop a chip for generative AI training.
The chips are designed for specific data tasks to improve energy efficiency and cost, and Meta is building entire liquid-cooled systems around them. This rapid in-house development is driven by Meta's massive data center expansion, with new chips planned at six-month intervals to support its infrastructure growth.
Main topics: Meta's in-house AI chip development (MTIA program), the focus on inference chips versus training chips, the drive for efficiency and cost savings, and the rapid expansion of data center infrastructure.
Meta Platforms on Wednesday unveiled a roadmap of four new chips that the company is making in-house, as it rapidly expands its data centers.
Like many big tech companies such as Alphabet and Microsoft, Meta has invested heavily in building a team that can design chips in-house in addition to purchasing off-the-shelf products made by Nvidia and Advanced Micro Devices.
Making chips designed to tackle the âspecific types of â data â crunching Meta requires can lead to designs that use less energy and at a better cost.
The new chips are part of the company's Meta Training and Inference Accelerator (MTIA) program and the first of the new chips called the MTIA 300 is in use powering the company's ranking and recommendation systems. The other three will be rolled out this year and in 2027, with the final two chips, the MTIA 450 and 500 being designed to perform â inference, the âprocess when an AI model such as the one that powers the ChatGPT app responds to customer queries and requests.
"We see inference demand exploding at â the moment and that's what we're currently focused on," Yee Jiun âSong, Meta's vice president of engineering, said in an interview.
Meta âhas had some success with inference chips but has struggled with its long-time ambitions to make a generative AI training chip, capable of building the large models that power AI apps.
Beginning with the MTIA 400, which the company says is on the path to being used in its data centers, Meta has designed an entire system around the chips, which âis roughly the size of several server racks and includes a version of liquid cooling.
The company plans to release the new chips at six-month intervals âbecause it is ârapidly expanding the number of â data centers it uses to run apps like Instagram and Facebook, Song said.
"That is the reality of how quickly our infrastructure is being built out," Song said.
The company said in âJanuary it expects capital spending of between $115 billion and $135 billion this year.
Meta contracts Broadcom to help with some elements of the designs, though Song did not specify which chips. The company uses Taiwan Semiconductor Manufacturing Co to fabricate the processors.
In February, Meta signed big deals with Nvidia and AMD to buy tens of billions of dollars worth of chips.
Like many big tech companies such as Alphabet and Microsoft, Meta has invested heavily in building a team that can design chips in-house in addition to purchasing off-the-shelf products made by Nvidia and Advanced Micro Devices.
Making chips designed to tackle the âspecific types of â data â crunching Meta requires can lead to designs that use less energy and at a better cost.
The new chips are part of the company's Meta Training and Inference Accelerator (MTIA) program and the first of the new chips called the MTIA 300 is in use powering the company's ranking and recommendation systems. The other three will be rolled out this year and in 2027, with the final two chips, the MTIA 450 and 500 being designed to perform â inference, the âprocess when an AI model such as the one that powers the ChatGPT app responds to customer queries and requests.
"We see inference demand exploding at â the moment and that's what we're currently focused on," Yee Jiun âSong, Meta's vice president of engineering, said in an interview.
Meta âhas had some success with inference chips but has struggled with its long-time ambitions to make a generative AI training chip, capable of building the large models that power AI apps.
Beginning with the MTIA 400, which the company says is on the path to being used in its data centers, Meta has designed an entire system around the chips, which âis roughly the size of several server racks and includes a version of liquid cooling.
The company plans to release the new chips at six-month intervals âbecause it is ârapidly expanding the number of â data centers it uses to run apps like Instagram and Facebook, Song said.
"That is the reality of how quickly our infrastructure is being built out," Song said.
The company said in âJanuary it expects capital spending of between $115 billion and $135 billion this year.
Meta contracts Broadcom to help with some elements of the designs, though Song did not specify which chips. The company uses Taiwan Semiconductor Manufacturing Co to fabricate the processors.
In February, Meta signed big deals with Nvidia and AMD to buy tens of billions of dollars worth of chips.
Meridiano
