Daimler chooses Xilinx for AI development

June 26, 2018 //By Christoph Hammerschmidt
Daimler chooses Xilinx for AI development
Luxury vehicle manufacturer Daimler cooperates with Xilinx in the development of vehicle applications based on Artificial Intelligence (AI). Initially, a technology collaboration with the technology provider is planned, focusing on Xilinx´ 16nm Zync Ultrascale+ MPSoC family.

As part of this strategic collaboration, deep learning experts from the Mercedes-Benz Research and Development Centers in Sindelfingen and Bangalore, India, are implementing their AI algorithms on a highly adaptable automotive platform from Xilinx. Mercedes-Benz will use the upcoming Xilinx AI processor technology to achieve the highest possible efficiency in the execution of its neural networks.

Daimler was quoted in a press release with a reference to "high energy efficiency under difficult thermal environmental conditions". According to information available to eeNews Europe, the efficiency of the SoC used did indeed play a decisive role in Daimler's choice in favour of Xilinx. Daimler also uses Nvdia processors, for example in the new AI-controlled In-Vehicle Information system MBUX in the Mercerdes A-class. However, developers often criticize the high energy demand of Nvidia chips.

Xilinx has a long track record in the automotive industry. In more than twelve years, the company has delivered more than 40 million automotive components to vehicle manufacturers and Tier 1 suppliers. The first cars with a neural network based on Xilinx chips will go into series production this year. Xilinx declined to name the manufacturer of this vehicle. However, industry observers consider it unlikely that it will be Daimler.

Related news:

GPU architecture not suited for AI, says Xilinx CEO

Continental relies on Xilinx expertise for auto-driving computer platform

Xilinx flexes its muscles in the ADAS arena


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.