Meta To Deploy in-house Custom Chips To Advance AI Drive By CIOTechOutlook Team

Meta To Deploy in-house Custom Chips To Advance AI Drive

CIOTechOutlook Team | Friday, 02 February 2024, 06:06 IST

Facebook owner Meta Stages intends to send into its server farms this year another rendition of a custom chip pointed toward supporting its artificial intelligence (AI) push, as indicated by an interior organization report.

The chip, a second generation of an in-house silicon line Meta announced last year, could help to reduce Meta's dependence on the Nvidia chips that dominate the market and control the spiraling costs associated with running AI workloads as it races to launch AI products, as per economic times.

The world's greatest online entertainment organization has been scrambling to support its registering limit with regards to the eager for power generative computer based intelligence items it is driving into applications Facebook, Instagram and WhatsApp and equipment gadgets like its Beam Boycott smartglasses, burning through billions of dollars to gather munititions stockpiles of specific chips and reconfigure server farms to oblige them.

According to Dylan Patel, the founder of the silicon research group SemiAnalysis, a successful deployment of Meta's own chip could potentially save billions of dollars in chip purchasing costs and hundreds of millions of dollars in annual energy costs.

The chips, framework and energy expected to run computer based intelligence applications has turned into a goliath sinkhole of speculation for tech organizations, somewhat balancing acquires made in the surge of fervor around the innovation.

A Meta representative affirmed the arrangement to place the refreshed chip into creation in 2024, saying it would work collaborating with the countless off-the-rack designs handling units (GPUs) - the go-to chips for AI based intelligence - the organization was purchasing.

"We see our internally developed accelerators to be highly complementary to commercially available GPUs in delivering the optimal mix of performance and efficiency on Meta-specific workloads," the spokesperson said in a statement.