Samsung uses artificial intelligence for the automated design of cutting-edge computers chips.
This South Korean company is among the first to employ AI in creating its chips. Samsung is using AI features in new software from Synopsys, a leading chip design software firm used by many companies. Aart De Geus (chairman and coCEO at Synopsys) says, “What you see here is the first real commercial processor design using AI.”
Nvidia and Google have also discussed AI-based chips. DSO.ai by Synopsys is the most powerful because it works with many companies. According to industry observers, the tool could accelerate semiconductor development and enable new chip designs.
Synopsys also has a valuable resource for creating AI-designed chips, years of semiconductor design that can be used in training an AI algorithm.
A spokesperson for Samsung confirms that the company is using Synopsys AI software to design its Exynos chips, which are used in smartphones, including its own branded handsets, as well as other gadgets. Samsung unveiled its newest smartphone, a foldable device called the Galaxy Z Fold3, earlier this week. It was not clear if the AI-designed chips are in production or what product they might appear in.
AI is changing how chips are manufactured across the board.
In June, Google published a research paper that described how it used AI to organize the Tensor chip components. It uses this to train and execute AI programs at its data center. Google’s next smartphone, the Pixel 6, will feature a custom chip manufactured by Samsung. An official at Google declined to confirm whether AI was involved in the design of the smartphone’s chip.
Nvidia, IBM and other chipmakers are also exploring AI-driven chips design. Cadence is a rival to Synopsys and another maker of chip-design software are developing AI tools that help with the creation of blueprints for new chips.
Mike Demler, a senior analyst at the Linley Group who tracks chip design software, says artificial intelligence is well suited to arranging billions of transistors across a chip. He says that artificial intelligence is well-suited to solving complex problems. It will become an integral part of any computational toolkit.
Demler said that AI is often expensive because it takes a large amount of computing power in the cloud to train an algorithm. He expects AI to be more affordable as computing costs drop and algorithms become more efficient. He also said that not all tasks in chip design can be automated. Expert designers will still be needed.
Microprocessors of today are extremely complex. They have many components and must be combined well. It takes a lot of time and decades to design a chip. Chip designers who are skilled in designing chips have a natural understanding of how decisions affect the process. That understanding cannot easily be written into computer code, but some of the same skill can be captured using machine learning.
Synopsys uses an AI technique, along with Google, Nvidia and IBM to design a chip. Reinforcement learning involves training an algorithm to perform a task through reward or punishment, and it has proven an effective way of capturing subtle and hard-to-codify human judgment.
This method automatically draws up basic design elements, such as the layout of the components, how they should be wired together. It also allows you to simulate different designs and learn which one produces the best results. The software can be used to speed up the design process and allows engineers to explore new designs faster. In a June blog post, Synopsys said one North American manufacturer of integrated circuits had improved the performance of a chip by 15 percent using the software.
DeepMind (a Google subsidiary) used reinforcement learning in 2016 to create AlphaGo. This program was capable of playing the game Go with enough skill to be able to beat a top-level Go player.
De Geus claims that his company discovered reinforcement learning was also useful in chip design. De Geus states that his company was able to achieve the same results in just one year and half as an expert team in just weeks. He will present details of the technology and its development at HotChips, a semiconductor technology conference, on August 23.
Stelios Diantis, Synopsys’ senior director for artificial intelligence solutions, said that the DSO.ai software could be set up to prioritize different goals such as energy efficiency or performance.
The semiconductors and the equipment used in making them have been increasingly valued. The US government has sought to restrict the supply of chipmaking technology to China, a key rival, and some politicians have called for software to be added to the export controls list.
AI-designed chips are also enabling developers to use AI to create software that runs more efficiently. This might include the neural network algorithms that run on specialized AI chips and are commonly used in modern AI.
“AI-powered codesign of software and hardware is a rapidly growing direction,” says Song Han, a professor at MIT who specializes in AI chip design. We have already seen some promising results.
- The latest on tech, science, and more: Get our newsletters!
- Prisoners, doctors, and the battle over trans medical care
- Scientists could one day float an aerial robot above Venus
- How to get the most out of your meditation app
- Covid is pulling the plug on beloved Japanese arcades
- China cracks down on its tech giants. Do you sound familiar?
- Explore AI like never before with our new database
- WIRED Games: Get the latest tips, reviews, and more
- Are things not working right? These are our top picks for wireless headphones, soundbars and Bluetooth speakers
Publited Fri, 13 August 2021 at 11:31:22 +0000