AI Designs Computer Chips We Can’t Understand — But They Work Really Well

Written by Mihai Andrei on . Posted in Current News

When chips are usually designed, scientists and engineers work with patterns and templates that are well-known.

A new study published in Nature Communications tried a different approach: a deep-learning-enabled design process for creating circuits and components.

Using artificial intelligence (AI), researchers at Princeton University and IIT Madras demonstrated an “inverse design” method, where you start from the desired properties and then make the design based on that.

The designs seem to work really well, but there’s a catch: no one really knows why they work so well.

“Humans cannot understand them, but they can work better,” said Kaushik Sengupta, the lead researcher, a professor of electrical and computer engineering at Princeton.

AI at the helm

The AI-driven method focused on designing wireless chips, which are extremely important for high-frequency applications like 5G networks, radar systems, and advanced sensing technologies.

These circuits power innovations in everything from radar systems to autonomous vehicles but their development is notoriously slow. Engineers would start on predefined templates and manually optimize or improve designs through iterative simulations and testing.

This method is time-consuming and challenging. It also requires a high degree of expertise, which limits just how much (and how fast) you can improve. This is where the new study comes in.

Whereas the previous method was bottom-up, the new approach is top-down. You start by thinking about what kind of properties you want and then figure out how you can do it.

The researchers trained convolutional neural networks (CNNs) — a type of AI model — to understand the complex relationship between a circuit’s geometry and its electromagnetic behavior. These models can predict how a proposed design will perform, often operating on a completely different type of design than what we’re used to.

The study showcased a range of use cases, from simple one-port antennas to complex multi-port RF (radio frequency) structures like filters, or power dividers. The AI-designed compact antennas that function across two distinct frequencies, improving performance for multi-band devices.

Within minutes, it synthesized filters with precise band-pass characteristics, a task that would have taken days or weeks before.

The ability to rapidly design high-performance circuits could accelerate advancements in telecommunications, autonomous systems, and beyond. This approach empowers engineers to focus on innovation rather than routine optimization. Yet perhaps the most exciting part is the new types of designs it came up with.

Completely new concepts

“We are coming up with structures that are complex and look randomly shaped, and when connected with circuits, they create previously unachievable performance,” says Sengupta. The designs were unintuitive and very different than those made by the human mind. Yet, they frequently offered significant improvements.

“Classical designs, carefully, put these circuits and electromagnetic elements together, piece by piece, so that the signal flows in the way we want it to flow in the chip. By changing those structures, we incorporate new properties,” Sengupta said. “Before, we had a finite way of doing this, but now the options are much larger.”

This study marks a pivotal moment in engineering, where AI not only accelerates innovation but also expands the boundaries of what’s possible.

Wireless chips are a combination of standard electronics (like computer chips) and electromagnetic components like antennas or signal splitters. While this study focuses on RF and sub-terahertz frequencies, the principles of AI-driven design can extend to computer chips or even quantum computing.

“There are pitfalls that still require human designers to correct,” Sengupta said. “The point is not to replace human designers with tools.

The point is to enhance productivity with new tools. The human mind is best utilized to create or invent new things, and the more mundane, utilitarian work can be offloaded to these tools.

But this also raises new questions.

The black box

How comfortable are we with these designs we don’t fully understand? What happens if something goes wrong with this design?

No doubt, AI will play an increasingly important role in how we design things, and chips are no exception. Yet for the most part, we still have no transparency on how AI arrives at its designs.

This makes it difficult for engineers to fully understand or predict the behavior of these circuits under all conditions. This “black-box” nature could lead to unforeseen failures or vulnerabilities, particularly in critical applications like medical devices, autonomous vehicles, or communication systems.

Additionally, if errors arise, tracing and rectifying the issue may prove more complex than in manually designed systems.

On a practical level, over-reliance on AI might erode the foundational knowledge and skills of human designers, creating a gap in expertise should the technology fail or be unavailable.

No doubt, we are heading towards a new age of design. Hopefully, it is one in which humans still hold the reins.

Journal Reference: Emir Ali Karahan et al, Deep-learning enabled generalized inverse design of multi-port radio-frequency and sub-terahertz passives and integrated circuits, Nature Communications (2024). DOI: 10.1038/s41467-024-54178-1

See more here zmescience.com

Please Donate Below To Support Our Ongoing Work To Defend The Scientific Method

PRINCIPIA SCIENTIFIC INTERNATIONAL, legally registered in the UK as a company incorporated for charitable purposes. Head Office: 27 Old Gloucester Street, London WC1N 3AX. 

Share via

Trackback from your site.

Comments (2)

  • Avatar

    Howdy

    |

    Microsoft has windows, that doesn’t work very well, but M$ can’t understand that either.

    “The designs seem to work really well, but there’s a catch: no one really knows why they work so well.”
    And that does not seem precarious?

    Reply

  • Avatar

    Tom

    |

    By the time these clowns figure out how they work, we will be enslaved by them with no way out. Can you imagine some A/i retards controlling the grid, or a nuclear arsenal, or some other military devices or airports and traffic?

    Reply

Leave a comment

Save my name, email, and website in this browser for the next time I comment.
Share via