Koç University develops photonic computing AI systems
Understand this event
On October 10, 2026, Tech Xplore reported that a team led by Koç University Assistant Professor Uğur Teğin had developed two systems for AI computing that use light rather than electronic circuits to perform computations. The report presented the work as a new photonic computing approach for sustainable AI, based on shifting some computational workload to light. The available material describes only the development of the two systems and provides no specific performance or energy consumption data, or information on application progress, so it is not yet possible to assess their practical impact.
AI 根据报道生成 · 4 days ago更新
Impact and actions
AI compilation of public coverage, separating facts, inference and scenarios. Verify rules against the original sources.
Photonics computing research shows potential for AI image classification; energy savings remain unverified
First update · Latest update (Beijing time)
Impact level
⚪ Background
⚠ Risk:Experimental classification accuracy cannot be directly extrapolated to general AI performance or commercial deployment outcomes.
What happened
On October 10, 2026, Tech Xplore reported on two photonic computing studies by a Koç University team. One system retained parallel processing of color channels and raised skin disease classification accuracy from 68% using raw color pixel data to 98%. Another used a multimode fiber laser to process data, achieving 85% to 99% accuracy across different image datasets. Both systems still rely on cameras and electronic control components. [1]
Involves:Programmable photonic neural network、Multimode fiber laser
Seller impact
- FactOptical systems can handle some feature processing, but the two studies use different tasks and baselines, so their accuracy figures cannot be used to compare performance. The materials provide no comparable figures for whole-system energy use, throughput, or cost versus electronic systems. [1]
- InferenceElectronic controls and optical integration may limit how much energy savings translate into the cost of real-world tools, so there is not enough evidence to support sellers switching AI tools. Revisit the papers over the next two weeks; commercial lead times and scalability are unknown. [1]
Suggested action
Check whether the two papers provide a whole-system energy comparison, including camera and control components, for the same task at the same accuracy.
Evidence that would disprove the assessment
- Assessment 1：If the papers’ supplementary materials provide comparisons of whole-system energy use, throughput, and cost for the same task at the same accuracy, the conclusion that comparable evidence is insufficient would no longer hold.
- Assessment 2：If a reproducible end-to-end test shows lower total energy use and operating costs for an equivalent task, and a purchasable solution is available, the assessment should be revisited.
Coverage timeline
Follow coverage to understand different perspectives.
- Tech Xplore: Technology, AI, and the Electronics IndustryOriginal ↗Shifting some computational workload to light: Koç University proposes a new photonic computing approach for sustainable AI
A team led by Koç University Assistant Professor Uğur Teğin developed two systems that use light rather than electronic circuits to perform AI computations.
Interest trend for this event
Current interest 3·Comparable scope peak 9（10/10 23:00）·Change over 24 hours within comparable scope -49%
Trends compare the same continuously observed entities, which may be fewer than the current interest count. Move the pointer or select the chart to see hourly values; use arrow keys to navigate.