Musk personally admitted: "Tesla is developing specialized artificial intelligence hardware, and will be the world's best hardware." The news about Tesla's own development of the chip is not the first time.
In September of this year, Sanjay Jha, the CEO of the world's second-largest wafer foundry Gemini (former GlobalFoundries), disclosed to the media at the Santa Clara grid core technology summit that Tesla was Work closely with foundries.
However, in response to this news, Gexin’s spokesperson clarified that he was misreading. The spokesperson said: "Gexin has not started substantive cooperation with Tesla. Sanjay Jia means that companies such as Apple, Google and Tesla cooperate with chip makers to provide differentiated services. Typical Case."
But now, Musk personally admitted: "Tesla is developing specialized artificial intelligence hardware and will be the world's best hardware."
According to foreign media reports, Tesla's founder and CEO Elon Musk announced for the first time at an event at the NIPS 2017 conference that Tesla is independently developing AI chips to improve the auto's ability to drive. At the same time, Musk also sent Twitter to predict that Tesla's first fully automatic driverless car will be implemented in two years.
Nvidia cannot meet Tesla’s needs and ambitions
As we all know, Tesla's self-driving cars have high standards for chips. At present, Tesla uses graphics processing chips provided by Nvidia, but this does not seem to satisfy Tesla's needs and ambitions.
Musk said: "Jim is developing specialized AI hardware. We think it will be the best hardware in the world. Dedicated hardware can one day be used in Tesla to do the calculations needed for automatic driving."
It is reported that Jim Keller (Jim Keller) is the object of worship of countless DIY players, his experience is legendary, Apple, AMD, Tesla and other large Silicon Valley companies have his presence, but he also led the design The famous AMDZEN architecture. Jim Keller currently serves as vice president of hardware for Tesla. He previously served as chip architect at AMD.
It is understood that Jim Keller will lead a 50-person team at Tesla to develop a dedicated AI chip for Tesla. This AI chip will be used in Tesla electric vehicles in the future to complete driverless driving. Operation operation. Tesla currently has more than 1.3 billion miles of autopilot mileage data, which are accumulated through the first-generation autopilot system. Jim Keller is likely to improve or continue to develop the next-generation autopilot system. chip.
When referring to the significance of the development of AI chips, Jim Keller said that custom chips can improve the computing efficiency and thus achieve better autopilot performance. Musk said that Tesla will achieve full self-driving within two years, when people can lie in the back seat of the car to sleep.
It is worth mentioning that just the same day that Musk announced Tesla is developing AI chips, Nvidia founder CEO Huang Renxun also appeared in Long Beach and announced the launch of a new GPU, the TitanV. TitanV's performance is 9 times that of the previous chip TitanXp, but the price is only slightly increased to 3,000 US dollars.
Increased competition in the smart chip market
In response to competition issues, Nvidia founder Huang Renxun has always maintained an open attitude. He told the First Financial reporter at the GPU Technology Conference in September this year: "Tesla is a very good car. I opened it because he had a strong sense of control, even if it was loaded with Intel chips. "Huang Renxun is a Tesla fan. He owns several Tesla.
In terms of driverless car chips, in addition to Nvidia, the strongest competitor is Google’s parent company, Alphabet. In addition, Apple also showed strong interest in driverless technology and actively developed automatic driving system technology. GM, Ford and other auto giants are also actively seeking cooperation with technology companies to jointly develop driverless technology.
A large number of artificial intelligence chip companies have also emerged in the domestic market, such as the Cambrian period. The Cambrian chip has already begun to be used on Huawei's smartphones and will release the Cambrian 1M processor for autonomous driving applications next year.
Although Tesla has not announced when it will sell this customized AI chip, this will undoubtedly make the entire artificial intelligence chip market more competitive.
Transparent Led Display,Transparent Led Screen,Transparent Led,Led Transparent Screen
ShenZhen Megagem Tech Co.,Ltd , https://www.megleddisplay.com