According to CNET, Google may have foreseen the spread of artificial intelligence (AI) technology in applications, devices, and services, such as identifying a friend's face in a photo and giving a smart speaker a human-like voice. The good news is that the processor industry has also noticed this. This means that we will see a large number of new chips that not only can accelerate AI tasks on mobile phones or laptops, but also can appear on car or home security cameras.
(Figure 1: Google's second-generation TPU AI chip. Google uses artificial intelligence to provide services such as search, translation, and face detection, but its AI chip (ie, TPU, tensor processing unit) runs in the data center. Now, AI chips are expanding into smart phones, personal computers, automobiles, etc.)
This is just the latest progress made by AI. This seemingly all-encompassing term is often referred to as what is known as "machine learning" or "neural network" technology, which is established by simulating the human brain. When you train neural networks with real-world data, it can find spam, transcribe your voice messages into text messages or identify cats and other objects. This has brought dramatic changes to the design of rigid traditional computer programs.
AI will become ubiquitous. For example, Facebook CEO Mark Zuckerberg had testified in Congress for nearly 10 hours and he mentioned many potential AI solutions for dealing with fake accounts and misinformation. At the same time, AI is also a popular vocabulary used by handset makers such as Apple, Samsung and LG.
Linley Gwennap, chief analyst at the chip analysis firm Linley Group, said at the processor conference: "Most new high-end smartphones have AI accelerators, including the Apple iPhone X A11 chip and the Samsung Galaxy S9 Exynos 9810. We have seen As time goes by, AI is gradually being applied to mid-range handsets, and it may continue to flow to low-end handsets over time."
AI chips may not only be used on smart phones, thanks to the startup's custom chip design and efforts from mobile chip giant ARM's Trillium AI project. This means that today's computer intelligence revolution is only just beginning. It can eventually help Amazon and Google's digital assistants expand to new devices, let your car recognize pedestrians and everything around them, make your computer smarter and automatically take photos And editing and so on.
Take a web security camera as an example. It can make your home network and broadband connection a continuous video stream. Gwennap explained: “The camera you really wanted was able to see the scene and then said nothing happened and there was no need to upload a video. When a change happened, it could send an image or a notification.” Gwennap anticipates that the PC can be The relatively powerful main processor runs AI, but as more software uses this technology, they also get AI chips. Adobe Systems software like Photoshop and Premiere Pro has pushed the processor to the limit, and the company has introduced Sensei AI technology to accelerate tasks like photo editing.
(Figure 2: Various sensors will be installed on driverless cars, and chip companies such as Synopsys and GlobalFoundries are expected to sell processors capable of handling all data)
As the driverless technology advances, the car will gain AI brain. The company called Synopsys showed its way at the conference to manage the large amount of data collected from radar, laser radar, cameras and other driverless cars. Synopsys marketing manager Gordon Cooper said: "Apple implanted neural network in iPhone X. Now everyone wants to have neural network function."
Today, smart speakers like Amazon Echo or Google Home are at the forefront of AI, but most of their “brains” rely on web links. Gwennap said that the built-in AI chips allow them to understand your voice commands faster, which is an important improvement, because no one likes to delay when ordering new music or turning the lights off.
In the field of drones, AI will mean more powerful capabilities, such as avoiding obstacles in flight, or tracking a moving object. For example, walk down a ski slope so that you can take action videos to the next level.
For the chip industry that has suffered from the rapid development of Moore’s Law, the addition of AI has brought a fresh air. Today, new laptops may not look much faster than they were five years ago, but with the advent of AI chips, you may have reason to upgrade again.
(Figure 1: Google's second-generation TPU AI chip. Google uses artificial intelligence to provide services such as search, translation, and face detection, but its AI chip (ie, TPU, tensor processing unit) runs in the data center. Now, AI chips are expanding into smart phones, personal computers, automobiles, etc.)
This is just the latest progress made by AI. This seemingly all-encompassing term is often referred to as what is known as "machine learning" or "neural network" technology, which is established by simulating the human brain. When you train neural networks with real-world data, it can find spam, transcribe your voice messages into text messages or identify cats and other objects. This has brought dramatic changes to the design of rigid traditional computer programs.
AI will become ubiquitous. For example, Facebook CEO Mark Zuckerberg had testified in Congress for nearly 10 hours and he mentioned many potential AI solutions for dealing with fake accounts and misinformation. At the same time, AI is also a popular vocabulary used by handset makers such as Apple, Samsung and LG.
Linley Gwennap, chief analyst at the chip analysis firm Linley Group, said at the processor conference: "Most new high-end smartphones have AI accelerators, including the Apple iPhone X A11 chip and the Samsung Galaxy S9 Exynos 9810. We have seen As time goes by, AI is gradually being applied to mid-range handsets, and it may continue to flow to low-end handsets over time."
AI chips may not only be used on smart phones, thanks to the startup's custom chip design and efforts from mobile chip giant ARM's Trillium AI project. This means that today's computer intelligence revolution is only just beginning. It can eventually help Amazon and Google's digital assistants expand to new devices, let your car recognize pedestrians and everything around them, make your computer smarter and automatically take photos And editing and so on.
Take a web security camera as an example. It can make your home network and broadband connection a continuous video stream. Gwennap explained: “The camera you really wanted was able to see the scene and then said nothing happened and there was no need to upload a video. When a change happened, it could send an image or a notification.” Gwennap anticipates that the PC can be The relatively powerful main processor runs AI, but as more software uses this technology, they also get AI chips. Adobe Systems software like Photoshop and Premiere Pro has pushed the processor to the limit, and the company has introduced Sensei AI technology to accelerate tasks like photo editing.
(Figure 2: Various sensors will be installed on driverless cars, and chip companies such as Synopsys and GlobalFoundries are expected to sell processors capable of handling all data)
As the driverless technology advances, the car will gain AI brain. The company called Synopsys showed its way at the conference to manage the large amount of data collected from radar, laser radar, cameras and other driverless cars. Synopsys marketing manager Gordon Cooper said: "Apple implanted neural network in iPhone X. Now everyone wants to have neural network function."
Today, smart speakers like Amazon Echo or Google Home are at the forefront of AI, but most of their “brains” rely on web links. Gwennap said that the built-in AI chips allow them to understand your voice commands faster, which is an important improvement, because no one likes to delay when ordering new music or turning the lights off.
In the field of drones, AI will mean more powerful capabilities, such as avoiding obstacles in flight, or tracking a moving object. For example, walk down a ski slope so that you can take action videos to the next level.
For the chip industry that has suffered from the rapid development of Moore’s Law, the addition of AI has brought a fresh air. Today, new laptops may not look much faster than they were five years ago, but with the advent of AI chips, you may have reason to upgrade again.
没有评论:
发表评论