AI chips have become king, and using AI to design, manufacture, and sell AI chips has become even more popular.
When giants such as Apple, NvIdia, Qualcomm, and Samsung launched their own AI chips, AI also became a powerful tool to accelerate the design, manufacturing, and even sales of AI chips. Now Intel is also actively using AI to design and manufacture the latest Meteor Lake processors, hoping that the new processors will turn laptops into AI PCs and regain the chip manufacturing leadership seized by TSMC.
For Intel, the most helpful aspect of AI tools is that they can detect defects in the early stages of the chip design process and speed up the launch of chips. AI tools can also help oversee chip manufacturing, allowing more Intel silicon to become part of products instead of being thrown into the trash.
Meteor Lake can be said to be a major processor with multiple "chiplets" stacked into a single package. Given the many nuances of each processor component, AI tools can help find the best way to sell each processor. Shlomit Weiss, co-leader of Intel's Design Engineering Group, said that with AI tools, the number of items that can be sold has increased significantly.
Use AI to ensure chip design compliance and find the best combination to sell chips
Intel is not the only one to see the benefits of AI tools. Cadence and Synopsys, two major processor design software suppliers, provide AI-enhanced tools for laying out processor components. Google has long announced the use of AI to develop its own AI acceleration processor. Nvidia, the most popular GPU manufacturer, uses AI to replace extremely slow traditional computing processes on a wider range of its production lines. Reinforcement learning can be said to be machine learning often used in chip design to explore various possible options and benefit when it is found to be closer to the goal.
Weiss and Navid Shahriari, another co-leader of the group, said Intel combines AI tools from outside chip design companies with its own tools. When simulating new designs, Intel AI analyzes and classifies defects to help engineers quickly debug them, saving hours of work every day.
Weiss said Intel is running OpenAI GPT generative AI tools modified to meet specific needs on its own systems to ensure chip designs comply with specifications. Shahriari said that since the functions of each production line chip will have slightly different clock speeds, caches and power consumption, Intel will use AI to carefully examine the chip and view the complex combination of related details to find the best combination of processor and commercial version. .
The latest Meteor Lake processors are made up of multiple small chips, or dies, which means there are more features to analyze, making it more complicated to figure out the best way to sell the chips. Fortunately, AI can find a balance to maximize output, efficiency and power improvements.
Related links: https://technews.tw/2023/12/19/ai-helps-chipmakers-design-the-very-processors-that-speed-up-ai/
【Disclaimer】
The content of this article only represents the author's personal views and has nothing to do with Kuiding.
The content, textual description and originality have not been confirmed by this website. This website does not make any guarantee or commitment for this article and all or part of its content, authenticity, completeness and timeliness. It is for readers' reference only. Please verify the relevant content by yourself.
馗鼎奈米科技股份有限公司 Creating Nano Technologies,Inc.
59 Alley 21 Lane 279, Chung Cheng Road, Yung Kang City, Tainan, TAIWAN
TEL: 886-6-2323927 FAX:886-6-2013306 URL: http://www.creating-nanotech.com