Facebook Hosts Casting Call For AI Accelerator ASIC And FPGA Chip Designers
While testifying before Congress over the Cambridge Analytical privacy scandal, Facebook CEO Mark Zuckerberg repeatedly talked about the need to further develop artificial intelligence (AI) technologies to tackle several of the challenges the company currently faces. Apparently that includes designing its own chips, too. Facebook is looking to hire silicon design engineers to develop ASIC and FPGA hardware that will drive its next generation technologies.
The job listing does not go into great detail about Facebook's semiconductor plans, though it does mention designing semi-custom and fully custom ASICs. By assembling its own chip design team, Facebook would ultimately make itself less reliant on companies like Intel and Qualcomm. It would also give Facebook the ability to design custom hardware specific to its AI and machine learning goals.
"The candidate would need to work with software and system engineers to understand limitations of current hardware and use their expertise to build custom solutions targeted at multiple verticals including AI/ML, compression, and video encoding. This position is full-time and located in our Menlo Park office," the job listing states.
There are a variety of use-case scenarios to consider. Bear in mind that
The bigger play, however, would likely center on Facebook's machine learning efforts. Facebook just doesn't have the people power to properly police every corner of the world's largest social network. On a daily basis the service contends with identifying and removing hate speech, fake news, spam posts, and everything else that goes against its terms of service. Much of that depends on users flagging posts as inappropriate, and then investigating those posts. But as Zuckerberg mentioned to Congress, that's a reactionary approach to a problem that requires a proactive effort.
Facebook is also involved in several outside projects, one of the biggest being trying to bring Internet connectivity to remote areas. Designing its own ASIC and FPGA hardware could play into those fields as well.
Top Image Source: Flickr via Marcin Wichary