US FPGA chips specialist Xilinx will step up development of self-adaptive AI chips for automotive applications based on the emerging concepts of Mobility as a Service (MaaS) and Transportation as a Service (TaaS) for next-generation, autonomous vehicles, according to industry sources.
Xilinx already shipped 170 million self-adaptive devices for automotive use in 2018, winning acclaims from clients, but further application of automotive AI chips will require the support of makers of diverse CMOS image sensors (CIS) and safety processors, including Sony and Infineon, the sources said.
Along with the rapid development of AI applications, common-use computing architectures can hardly satisfy the applications. But Xilinx's self-adaptive FPGA chips can, through the OTA (over the air) format, serve as a multi-core heterogeneous computing platform to help achieve upgrades of hardware components, the sources continued.
Xilinx has maintained close partnerships with TSMC, contracting the foundry to fabricate many advanced adaptive FPGA chips solutions with 7nm node for national defense, aeronautic and other applications, the sources added.
Article translated by Willis Ke
