TikTok Owner ByteDance Will Design Its Own Server SoCs
ByteDance, the owner of the TikTok social media app, has begun hiring professionals to develop system-on-chip (SoC) data centers. ByteDance will join several high-tech Chinese companies that have recently deployed their own data center SoCs designed to operate on specific workloads.
So far, ByteDance reports that it has posted 31 chip-related openings on its website. TMTPost.. While 31 chip engineers can hardly build advanced processors, they can create a set of design requirements for contract chip developers such as Alchip. Then, when the contract chip developer creates the appropriate design, the semiconductor contract manufacturer (such as TSMC) mass-produces it for ByteDance.
So far, ByteDance just says it plans to develop the chip without giving details. The chip design cycle usually takes years, so after a few years you will only learn about your company’s SoC.
What exactly ByteDance plans to develop is a big issue. CNBC (Opens in a new tab) A spokesperson for the company said that data center-grade system-on-chips by social media giants will support a variety of business areas such as video platforms, information and entertainment apps. Meanwhile, some sources based in China (and others) have shown that ByteDance is looking forward to designing artificial intelligence chips.
Since ByteDance’s main business is the TikTok media app, it makes sense to develop an SoC similar to Google’s Argos Video (Trans) Coding (VCU) to transcode video streams uploaded by TikTok users. .. Meanwhile, to further reduce the power consumption, traffic, and storage capacity used in the TikTok data center, ByteDance may introduce entirely new codecs or AI algorithms. In addition, adding AI extensions may improve image quality.
ByteDance will be another China-based cloud giant developing chips. For example, Alibaba has its own AI inference chip and general purpose processor. Baidu, on the other hand, has its own Kunlun AI processor. Tencent also has processors for a variety of cloud workloads.