Everyone is familiar with the chip, often in American science fiction movies, they will see that scientists often put the chip into the robot's brain, and then the robot can have all the same ability as real people. The chip is just like the human brain, carrying all the thoughts and actions. If a person has no brain, it is like a dead person. A smart machine has no chips, and it becomes a waste of copper.
In the age of artificial intelligence, chips are even more important. The smart cars, autopilots, and robots that were previously only seen on TV are no longer science fiction. They began to take place slowly and become a reality.
Google’s AlphaGo man-machine battle, Robin Li riding on a driverless car to Beijing’s Wuhuan, has actually happened to us. The source of all this is the need for a powerful brain based on the deep learning of artificial intelligence. ) to carry everything.
As the core giants of artificial intelligence will naturally not give up this piece of fat, so around the automatic driving, launched an AI chip battle for war.
With the chip's old-time player Intel, NVIDIA, which supports the game's visual GPU, and Google, the most autonomous Google of autopilot technology, it has formed a three-legged situation. Who will become the ultimate winner of the AI ​​chip has also become a concern in the industry.
NVIDIA has found that GPUs are used in artificial intelligence and quickly accelerate the pace of entry into the autopilot. They have already partnered with Tesla and Volvo.
NVIDIA, a fabless IC semiconductor company focused on designing smart core chipsets, initially designed gaming chips and later became famous for making GPUs that support game vision.
After years of development, NVIDIA has found that GPUs can also be applied to artificial intelligence. Artificial intelligence has only just begun to emerge in Silicon Valley. After seeing the opportunity, it immediately entered the auto-pilot, and the pace was rapid.
In a sense, NVIDIA promoted the development of self-driving cars and brought a powerful boost to automotive companies and research institutes. Many companies, with NVIDIA's DGX-1, have achieved full integration with hardware, in-depth learning software and development tools. In short, it is speeding up the design of more complex neural networks for deep learning.
Since NVIDIA entered the auto-piloting technology, it has continuously researched the open artificial intelligence vehicle computing platform and in 2015 launched the world's first Car PX, the first generation super brain. With the continuous updating of technology, the advantages are getting larger and larger. NVIDIA has become a leading company in autopilot hardware.
Currently NVIDIA has partnered with more than 70 well-known car companies, including some transportation network providers and autonomous driving technology companies. Not long ago, Volvo joined the NVIDIA camp, and the famous electric car manufacturer Tesla adopted the Drive PX 2 autopilot chip with Autopilot 2.0.
In catching up with the AI, NVIDIA's market value has skyrocketed in the past year. After that, Japan’s Softbank’s increasing holdings of stocks, Amazon’s and Microsoft’s cloud computing support, all seem to feel that NVIDIA is in the category. Winners.
Intel, the biggest rival, seems to have no recourse.
The author believes that it seems that all this is designed as well, as NVIDIA's biggest competitor, Intel, seems to have no recourse.
Intel, as the world's largest semiconductor chip manufacturer, has dominated the entire Internet market in the era of PC Internet and mobile Internet, and even IBM, which is most likely to match it, cannot parry.
In the battle for artificial intelligence and droneless chips, Intel has been able to meet the special needs of artificial intelligence chips, and has acquired a large number of related companies and powerful companies to develop custom chips to counter NVIDIA's challenges. NVIDIA's gamble.
In the age of artificial intelligence, subdivision seems to be important, as is the core chip for creating artificial intelligence. As a result, a rich variety of chips is also being bred, and human dependence on the central processor seems to be less serious than before. Intel's Xeon Phi, which supports AI-computing processors, will also take some time to formally market.
In my opinion, Intel's pace seems to be far behind NVIDIA.
We also seem to forget a heavyweight player, search giant Google. Google is a world-renowned giant. It can be seen everywhere in the field of science and technology. From drones to high-speed networks, it is involved in almost every area. It also makes Google take the initiative in autonomous driving technology. The forefront of the world.
Although Google is the best in autopilot, it is indeed in the chip field that it later invaded, and Google previously used NVIDIA's chips. Although Google has introduced a deep-learning TPU chip, it is still impossible to compete with NVIDIA and Intel.
Then, who is the ultimate winner of the artificial intelligence era AI chip, I believe that in the continuous improvement of deep learning technology, the future of artificial intelligence machine thrust capability will be more and more perfect, maybe one day will be able to meet all The ultimate algorithm comes out, half way out to replace NVIDIA.
Furthermore, from the perspective of time cost and capital cost, NVIDIA is totally unable to compare with Google. With Google's strength, once all aspects mature, it is entirely possible to replace NVIDIA and seize the dominance of AI chips. Who will become the ultimate player, we can only wait and see.
British Switch Socket W80 Series
Fused Socket,Remote Control Plug Socket,Modular Switch Socket,Chrome Switches And Sockets
Guangdong Jinli Electrical Appliance Co., Ltd , https://www.sokelectric.com