Researchers at MIT have developed a microscale porous silicon chip that solves matrix equations using waste heat instead of electricity.
Scientists in the US have created a tiny silicon chip that can perform mathematical operations in an electronic device using waste heat , and could transform energy-efficient thermal sensing and signal processing.
The device, built by Giuseppe Romano, PhD, a researcher at the Massachusetts Institute of Technology’s Institute for Soldier Nanotechnologies, and his team, introduces a novel analog computing technique where waste heat is the medium for data processing. Rather than relying on digital bits or voltage-based logic, it encodes input data as precise temperatures. These thermal inputs then flow through specially designed porous silicon structures. The resulting heat distribution represents the output of the computation.According to the researchers, the final result of the computation is represented by the amount of thermal energy collected at the output of the structure, which is maintained at a fixed temperature.Rethinking wasted energyThe team used the structures to demonstrate matrix vector multiplication with over 99 percent accuracy. Matrix multiplication is a fundamental mathematical technique machine-learning models like large language models utilize to process data and make predictions.Even though scaling the method for modern deep-learning models still remains a challenge, it could be used to detect heat sources and track temperature changes in electronics, at no additional energy cost. It could also replace multiple on-chip temperature sensors.“Most of the time, when you are performing computations in an electronic device, heat is the waste product,” Caio Silva, an undergraduate physics student at MIT, and lead author of the study, stated. “You often want to get rid of as much heat as you can,” Silva added. He said the team took the opposite approach, treating heat itself as a carrier of information and demonstrating that computation using heat is possible.For the project, the team developed a new software system that allowed them to design a material that can conduct heat in a specific manner. It uses a technique called inverse design that works backward. The scientists first define the desired function, then use algorithms to iteratively design the best geometry for the task.Using the system, the team designed complex silicon structures. Each is roughly the size of a dust particle, and can carry out computations using heat conduction, a form of analog computing that encodes and processes data using continuous values.From heat to mathThe researchers input a matrix defining the calculation. Using a grid, the software designs porous rectangular silicon structures. It then adjusts each pixel until it achieves the desired function. Meanwhile, heat flowing through the silicon carries out the matrix multiplication, with the structure’s shape encoding the coefficients.“These structures are far too complicated for us to come up with just through our own intuition,” Romano said. “We need to teach a computer to design them for us. That is what makes inverse design a very powerful technique.”However, because of heat conduction the structures could only encode positive coefficients. The scientists solved the challenge by splitting the target matrix into positive and negative components. They then processed them separately, and combined the results afterward.Using simulation, the team then tested the structures on simple matrices with two or three columns. Though simple these matrices are relevant for applications such as fusion sensing and microelectronics. The structures performed computations with more than 99 percent accuracy in many cases.But, scaling the method to large applications like deep learning still remains challenging, as it would require tiling millions of structures together. As the matrices become more complicated, the structures become less accurate.But because the structures rely on excess heat, they could be directly applied for tasks like thermal management, as well as heat source or temperature gradient detection in microelectronics. “If we have a localized heat source where we don’t want a heat source, it means we have a problem,” Romano concluded in a press release. “We could directly detect such heat sources with these structures, and we can just plug them in without needing any digital components.”The study has been published in the journal Physical Review Applied.
Environment Heat Innovation Inventions And Machines MIT Physics Silicon Chips Sustainability USA Waste Heat
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Shiba Inu Burn Rate Slumps 99%, Will Price Recovery Stall?Shiba Inu burn metric slows down despite ongoing price recovery, stirring doubts about a sustainable price rally.
Read more »
Cavs spoil LeBron James’ teary-eyed homecoming in 129-99 blowout win over LakersThe soaring Cavs spoiled LeBron James’ highly anticipated visit Wednesday night, overwhelming the Los Angeles Lakers, 129-99, in front of a sellout crowd at Rocket Arena.
Read more »
Cavaliers spoil LeBron James' return, riding big 3rd quarter to 129-99 rompDonovan Mitchell scored 25 points, Jaylon Tyson had 20 and the Cleveland Cavaliers used a big third quarter to beat the Los Angeles Lakers 129-99 and spoil LeBron James’ return to Northeast Ohio
Read more »
Cavaliers spoil LeBron James' return, riding big 3rd quarter to 129-99 rompDonovan Mitchell scored 25 points, Jaylon Tyson had 20 and the Cleveland Cavaliers used a big third quarter to beat the Los Angeles Lakers 129-99 and spoil LeBron James’ return to Northeast Ohio
Read more »
Cavaliers spoil LeBron James' return, riding big 3rd quarter to 129-99 romp- Donovan Mitchell scored 25 points, Jaylon Tyson had 20 and the Cleveland Cavaliers used a big third quarter to beat the Los Angeles Lakers 129-99 and spoil LeBron James' return to Northeast Ohio.
Read more »
Cavaliers spoil LeBron James' return, riding big 3rd quarter to 129-99 rompDonovan Mitchell scored 25 points, Jaylon Tyson had 20 and the Cleveland Cavaliers used a big third quarter to beat the Los Angeles Lakers 129-99 and spoil LeBron James' return to Northeast Ohio.
Read more »