The realm of processor manufacturing, dominated by giants like Apple, Nvidia, Qualcomm, Samsung, and others, is witnessing a transformative shift.
AI, long a beneficiary of advanced chip design, is now pivotal in creating the processors it runs on. A prime example of this evolution is Intel’s approach to its new Meteor Lake processors.
Intel leverages AI for Meteor Lake processor development
Intel, headquartered in Santa Clara, California, plans to unveil details about these processors, intended to redefine laptops as “AI PCs.” Notably, Intel has harnessed AI in developing these processors, representing a significant change in the landscape of chip design.
AI tools are revolutionizing chip manufacturing by catching bugs early in the design process, accelerating market readiness. They also optimize manufacturing, reducing waste and increasing the number of usable silicon slices.
For Intel’s Meteor Lake, the first significant processor comprised of multiple chiplets in a single package, AI has been instrumental in determining the most effective way to sell each processor, considering the slight variations in components.
Shlomit Weiss, co-leader of Intel’s Design Engineering Group, emphasized this in an interview, stating, “The amount of sellable units is increasing significantly.”
Read More: Billionaire Bill Ackman Accuses Harvard of DEI-Driven Hiring
Intel’s AI push: Advancing chipmaking, boosting performance
This shift is crucial for Intel as it strives to regain its leadership in chipmaking and fortify the U.S. industrial base. The implications extend to millions of Windows PC users, who stand to benefit from improved laptop performance and cost-effectiveness.
Modern processors, composed of billions of transistors, require sophisticated arrangements far beyond human capabilities. AI’s introduction into this field marks a significant advancement. These systems, adept at pattern recognition in complex data, are now aiding humans in refining processor production.
David Kanter, an analyst at Real World Insights, notes the necessity of this evolution: “As Moore’s Law progresses and we get more and more transistors, you can’t hand-tune everything anymore.”
Widespread adoption of AI in processor design across Tech giants
Intel’s use of AI in processor design is not an isolated case. Companies like Cadence and Synopsys, prominent in processor design software, are incorporating AI-enhanced tools.
Google and Nvidia have also utilized AI in developing their specialized processors, replacing slower traditional methods.
One AI approach, known as reinforcement learning, involves an AI system exploring options and being rewarded for outcomes that align with desired goals.
Elyse Rosenbaum, a professor of engineering at the University of Illinois, explains the potential of this method to reduce human involvement in the process.
Also Read: Faith and Politics: Presidential Hopefuls Share Their Views at Iowa University as 2024 looms large
Intel’s blend of external and custom AI tools for enhanced chip design
Intel combines external AI tools with its in-house developments, as explained by Weiss and her co-leader, Navid Shahriari.
Their internal tools include AI that categorizes bugs during simulations, allowing engineers to address issues more efficiently. “It can save a few hours for several people per day,” Weiss noted.
Additionally, Intel is experimenting with generative AI tools to interpret detailed chip specifications and create tests to ensure designs meet these specifications. Utilizing OpenAI’s GPT, adapted for Intel’s needs, the company is enhancing accuracy and reducing human error.
AI enhances Intel’s Meteor Lake chip selection and efficiency
AI’s application extends to scrutinizing chips off the production line, each with unique characteristics like clock speed and power consumption.
The AI system evaluates these details to optimize categorizing processors into various marketable versions.
Meteor Lake’s multiple interconnected chiplets pose a more significant challenge, but AI successfully navigates this complexity.
Shahriari highlights this, saying, “You can find that sweet spot of the multiple die that’s on that one Meteor Lake device.” The result is maximized yield and improved performance and power efficiency.
Read Next: President Biden Kicks Off 2024 Hollywood Fundraising Blitz Amidst Growing Popularity Concerns
