Meta and Google have announced the introduction of home-grown processors for their data centres. At the same time, Microsoft has declared it will invest billions into AI and cloud infrastructure in Japan.
This follows increased demand for AI services and a need to increase computing power.
Meta CEO Mark Zuckerburg said in January that the company planned to purchase 350,000 H100 processors from Nvidia.
Nvidia is currently the world’s leading chip producer, and the H100 chip, its previous flagship, has an estimated value of R562,000 each. That’s a total of around R196.7 billion.
Meta unveiled its own processor, the Meta Training and Inference Accelerator (MTIA), on Wednesday to power its AI services.
This was seen as an attempt to decrease its reliance on Nvidia chips.
Google and Microsft have also announced similar ventures within the past week.
Google releases Axion
Google announced the release of a general compute chip, Axion, to use in its data centres.
It will be the company’s first custom Arm-based CPU.
Google maintains that although compute accelerators transform consumers’ capabilities, general-purpose computing remains a significant part of customer workloads.
“Amdahl’s Law suggests that as accelerators continue to improve, general purpose compute will dominate the cost and limit the capability of our infrastructure unless we make commensurate investments to keep up,” Google said in a press release.
“Axion processors combine Google’s silicon expertise with Arm’s highest performing CPU cores to deliver instances with up to 30% better performance than the fastest general-purpose Arm-based instances available in the cloud today”, it added.
“[They deliver] up to 50% better performance and up to 60% better energy-efficiency than comparable current-generation x86-based instances.”
This is not Google’s first processor, as it released the Tensor Processing Unit in 2015 and the Video Coding Unit (VCU) in 2018, followed by the first of three generations of Tensor chips for smartphones in 2021.
Microsoft invests in Japan
Microsoft announced that it would invest $2.9 billion (R54.3 billion) in Japan’s cloud computing and AI infrastructure over the next two years.
This investment also focuses on providing AI skilling to more than three million people over three years.
“This significant enhancement in digital capacity will enable Microsoft to provide more advanced computing resources in Japan, including the latest graphics processing units (GPUs), which are crucial for speeding up AI workloads,” a Microsoft press release said.
Microsft will also open a Microsoft Research Asia lab in Tokyo, “showing long-term commitment to Japan.”
“The new lab will have a unique focus on areas including embodied AI and robotics, societal AI and wellbeing, and scientific discovery that align with Japan’s socioeconomic priorities.”
Samsung invests $44 billion
In related chip manufacturing news, Samsung will be investing $44 billion (R824 billion) into US chipmaking.
$6 billion of this was granted by the US government as part of The Chips and Science Act, which set aside $39 billion as part of a broader effort to bring processor production back to the US.
In 2021, Samsung began its investment with a $17 billion factory in Taylor, Texas. It is expected to create 1,800 jobs. Construction started in 2022 and production is supposed to start this year.
Samsung is set to announce where the remaining $27 billion will be invested soon, according to a Bloomberg report.