As traditional chips near physical limits and AI demand grows, energy use from AI data centers is projected to double by the end of the decade. This projection raises urgent questions about how to design more sustainable computing systems.
Researchers are studying neuromorphic computing, an approach that models hardware on biological neural networks. Suchi Guha, a professor of physics at the University of Missouri, notes that the brain is highly efficient, performing complex tasks on about 20 watts. Guha is also a core faculty member with the Materials Science and Engineering Institute.
The research team builds electronic components that mimic synapses. These organic transistors are designed to store and process information in the same place, rather than moving data back and forth between separate memory and processing units. That data movement both slows performance and consumes energy, so combining functions could reduce the cost.
In tests the scientists compared organic materials that looked similar but performed very differently when made into synaptic transistors. They found the key factor was the interface where the semiconductor meets an insulating layer. The study appears in ACS Applied Electronic Materials and lists coauthors from Mizzou and Hamad Bin Khalifa University. Source: University of Missouri.
Difficult words
- projection — a prediction about future events or amounts
- sustainable — able to continue without harming the environment
- neuromorphic — computing that copies biological neural network structure
- synapse — the connection point between two nerve cellssynapses
- transistor — an electronic switch that controls electrical signalstransistors
- interface — a surface or place where two things meet
- semiconductor — a material that partly conducts electric current
- insulating — not allowing heat or electricity to pass through
Tip: hover, focus or tap highlighted words in the article to see quick definitions while you read or listen.
Discussion questions
- Do you think designing computers like the brain is a good idea? Why or why not?
- How might lower energy use in AI data centers affect your daily life or community?
- What practical problems could scientists face when they combine memory and processing in one device?
Related articles
New training method helps models do long multiplication
Researchers studied why modern language models fail at long multiplication and compared standard fine-tuning with an Implicit Chain of Thought (ICoT) method. ICoT models learned to store intermediate results and reached perfect accuracy.
Zenica School of Comics: Art and Education for Children
The Zenica School of Comics began during the 1992–95 war and has taught around 200 young artists. The school still runs, faces changes from tablets and AI, and the regional comics scene survives through festivals and cooperation.
Touchscreens on car dashboards increase driver distraction
A simulator study found that using a car touchscreen while driving makes steering and touchscreen tasks worse. Multitasking reduced lane control and touchscreen accuracy; researchers suggest simple sensors could monitor attention and change the interface.
New ultrasound method improves diagnosis of breast masses
Researchers developed a new ultrasound signal-processing method that distinguishes fluid cysts from solid breast masses. In initial patient tests doctors identified masses correctly far more often than with conventional ultrasound, which could reduce biopsies and follow-ups.