Facebook Developing Machine Learning Chip – Report

Facebook - Shutterstock - © Pan Xunbin / Shutterstock.com

Engineers at Facebook are reportedly developing a machine learning chip, as part of its internal chip development program

Facebook continues to advance its inhouse chip development program, after a media report suggested the social networking giant is currently developing a machine learning chip.

According to a report in the Information, citing two people familiar with the project, the machine learning chip being developed by the social networking giant is designed to handle tasks such as content recommendation to users.

Facebook has been developing its own chips for a while now, in order to wean itself off its dependency on Intel.

In 2018 it was reported that Facebook was building an internal team to design its own semiconductors.

Machine learning chip

Facebook will utilise any inhouse developed processors in its servers, and to power artificial intelligence systems located within its data centres.

By designing its own silicon, the thinking is that Facebook gains finer control over product development and is be able to better tune its software and hardware together.

According to the report in the Information, besides the machine learning chip, Facebook has also developed another chip for video transcoding to improve the experience of watching recorded and live-streamed videos on its apps.

In 2019, Facebook said in a blog that it was developing custom chip designs specially meant to handle AI inference and video transcoding to improve performance, power and efficiency of its infrastructure.

The social networking giant also said it would work with semiconductor players such as Qualcomm, Intel and Marvell Technology to build these custom chips as general-purpose processors.

Yet The Information’s report suggests that Facebook is designing these chips completely in-house and without the assistance of these firms.

“Facebook is always exploring ways to drive greater levels of compute performance and power efficiency with our silicon partners and through our own internal efforts,” a company spokesperson reportedly said.

Other firms

Facebook is not alone in developing in-house processors.

Google is known to have developed its own artificial intelligence chip, as well as a chip for quantum computing.

Amazon Web Services has already designed a second generation data centre processor that is said to be more powerful, with at least a 20 percent performance increase over the first generation chip.

Microsoft is also developing its own ARM-based chips for its Azure servers, and even possible future Surface devices.

Apple made a very public break with Intel and instead increasingly utilises its own in-house central processing units (CPUs).

And lets not forget that Facebook also makes consumer facing hardware, in the form of the Oculus virtual reality headsets, most of which currently utilise Qualcomm processors.