Meta tests its first dedicated chip for artificial intelligence training

Meta, a social network giant, is testing its first dedicated chip for artificial intelligence training. This is part of the company’s long -term plan to reduce staggering infrastructure costs and reduce dependence on suppliers such as Nvidia.

According to Reuters, Meta has recently designed a new chip and tested it on a small scale. If this test is successful, the production of this chip begins on a large scale. The main purpose of this chip is to optimize the process of training artificial intelligence models and reduce energy consumption compared to conventional graphics processors that are currently used in this area.

Located in the Meta Training and Inference Accelerator (MTIA) series, the chip will be first used in meta content offer systems, such as Instagram and Facebook. Next, Meta intends to use this processor to teach productive artificial intelligence models, such as the Meta AI chat.

One of the benefits of this chip compared to traditional graphics processors is optimal energy consumption and higher processing power in specific artificial intelligence tasks. This can significantly reduce meta costs in the area of ​​artificial intelligence infrastructure.

Meta competition with Nvidia in the field of artificial intelligence chips

Meta has been one of Nvidia’s largest customers and has spent billions of dollars on buying graphic processors to train its artificial intelligence models, but Meta’s dedicated chip test shows that the company’s strategy has changed and moves towards independence in the field of artificial intelligence hardware.

According to experts, if the meta chip can perform successful performance in teaching artificial intelligence models, the company could compete with Nvidia and other artificial intelligence chips in the future. At the same time, it is still unclear whether the meta can reduce its dependence on Nvidia.

With the development of its first dedicated artificial intelligence chip, Meta has taken a big step towards reducing dependence on foreign suppliers and reducing infrastructure costs. If the chip can succeed in the experiments, Meta will become one of the major actors in the artificial hardware industry.