It is likely common knowledge that companies like Apple, Nvidia, Qualcomm, Samsung, and nearly all chip manufacturers are incorporating AI task acceleration circuitry into their products. You might be surprised to hear that artificial intelligence, through processes like AI Helping In Designing Chips, is also being utilized to speed up the creation of those chips. Let’s explore more of this inter-twining technology in this blog.
Intel is one example. On Thursday, the Santa Clara, California-based business unvieled details of its new Meteor Lake chips, which it claims will turn laptops into “AI PCs.” However, Intel used AI as well to make those chips.
The Process Of AI Helping In Designing Chips
AI tools monitor manufacturing to ensure that more of Intel’s tiny silicon slices end up inside devices rather than in the trash can, and they identify faults early in the design phase to speed up the release of chips. Given the numerous, minute variations in each processor’s componentry, artificial intelligence technologies assist in determining the optimal strategy for marketing Intel’s first large processor, Meteor Lake, which consists of several “chiplets” packed into a single container.
As a result of these AI technologies, “the amount of sellable units is increasing significantly,” as co-leader of Intel’s Design Engineering Group Shlomit Weiss stated in a special interview.
That is essential for Intel, which is rushing to restore the US industrial base and regain the supremacy in chipmaking that it lost to Taiwan Semiconductor Manufacturing Co. (TSMC). Furthermore, it matters to you if you’re one of the millions of Windows PC users out there since improved manufacturing translates into more affordable and more capable laptops.
These days, processors are made up of tens of billions of transistors, which are tiny electronic on-off switches. Through them, information in the form of ones and zeros flows to compute your income tax, spell-check your resume, and improve your phone photo. Since humans are incapable of arranging all those transistors in the best possible way or verifying that they will function as needed, computers have long been a necessary tool in the processor manufacturing industry.
Experts View
But applying AI is a novel approach. Artificial intelligence systems are capable of tasks like spam filtering and face recognition from your phone’s camera roll because they are trained to spot patterns in complex real-world data. They are now assisting humans in better managing the manufacturing of processors.
Real World Insights analyst David Kanter stated, “You can’t hand-tune everything anymore as Moore’s Law progresses and we get more and more transistors.”
It’s not just Intel. AI-enhanced tools for arranging processor pieces are available from Cadence and Synopsys, two significant manufacturers of the software tools used in CPU design. Google has detailed the technique by which it developed its AI acceleration processors using AI. Furthermore, Nvidia, the leading provider of AI and graphics chips, has replaced a lot of the slower traditional processing processes in the creation of its goods with AI.
The Process Of Reinforcement Learning
Reinforcement learning is one kind of AI that is utilized in the system. In it, an AI system looks at many possibilities and gets rewarded for identifying solutions that get closer to the intended result. “That reinforcement learning can potentially take the human out of that loop,” University of Illinois engineering professor Elyse Rosenbaum stated.
Read | 5 Best Antivirus Software
Navid Shahriari’s Take On AI Helping In Designing Chips
As per Weiss and her co-leader, Navid Shahriari, Intel is fusing AI technologies from outside chip design businesses with its own in-house produced capabilities. Here are a few instances of internal tools:
- Intel AI examines bug reports and classifies them while modeling new designs, a process that enables human engineers to begin problem-solving sooner. “It can save a few hours for several people per day,” Weiss stated.
- In order to produce tests that genuinely guarantee chip designs satisfy specifications, Intel has started testing generative AI tools that analyze written specifications in great detail. Today, people still perform the labor, but Weiss noted that errors can happen. Intel is utilizing OpenAI’s GPT, a massive language model that runs internally on Intel’s systems and has been customized for Intel’s specific requirements.
- With each chip coming off the assembly line having a slightly varied clock speed, cache memory size, and power consumption, Intel employs artificial intelligence (AI) to examine these chips closely. According to Shahriari, the AI system can determine the optimal method to slot chips into one of the many versions that Intel will sell by looking at this intricate collection of connected facts.
Since Meteor Lake has several interconnected chiplets, also known as die, there are many more combinations of all those features to consider when determining the optimal way to sell a chip.
However, the AI system can strike the correct balance. And that can be done by increasing the yield—that is, the percentage of chips that are useable.
“You can find that sweet spot of the multiple die that’s on that one Meteor Lake device,” Shahriari stated. “It not only maximizes the yield but also gives you performance and power improvement.,” she further said.
Wrapping Up
As AI seamlessly intertwines with the creation of processors, fostering a recursive loop of innovation, the horizon of possibilities expands through intelligent collaborations. This synergy, central to AI helping in designing chips, redefines the future of both AI and chip technology. In this ever-evolving landscape, the partnership propels us into an era where innovation knows no bounds. Stay tuned to GTA4.in for more insights into the fascinating world of cutting-edge technology.
Leave a Reply