The contribution of light chip could drastically reduce the energy consumption of AI first appeared in the online magazine Basic Thinking. You can start the day well every morning via our newsletter update.
AI chips are becoming increasingly powerful. Light instead of electricity could now be the next quantum leap: researchers have developed a light chip that could significantly reduce the energy consumption of artificial intelligence.
Artificial intelligence is hungry: not only for data, but also according to energy. The enormous electricity consumption of AI data centers has long since become a serious problem. But a new AI chip could help.
Because: A team of engineers has a computer chip developedthat does not rely on electricity, but on light. The result: up to 100 times higher energy efficiency-without loss of performance. So far, AI chips have been based on a process called “Convolution”, with which AI systems can interpret photos, videos or texts and recognize patterns.
So far, this calculation has been extremely resource -intensive and time -consuming. The new light chip bypasses this process because it has tiny lentils and lasers directly on the board. This allows him to carry out calculations with significantly less energy and in a shorter time.
With light instead of electricity: New chip increases efficiency from AI
The chip also no longer sends data via electrical lines, but via light rays. In the first tests, a prototype achieved an accuracy of 98 percent when recognizing handwritten digits and thus achieving the performance of conventional electronic chips.
The researchers achieved efficiency that is ten to 100 times higher than that of current chips that carry out the same calculations. “Exact an important mechanical learning process with almost zero energy is a quantum leap for future AI systems,” says study director Volker J. Sorger.
The technology behind the light chip are so-called fresnel lenses in miniature format. These two -dimensional versions of the same lenses that can be found in lighthouses are only a fraction of human hair. Data, for example from an image, are converted directly into laser light on the chip and guided through these lenses.
The result is then converted back into a digital signal in order to complete the task of the AI. The use of light has other advantages, because the system can process several data flows at the same time by using different light colors.
Brilliant future for AI?
According to the researchers, it is the first time that this type of optical calculation was applied on a chip and integrated into a AI neuronal network. This development could significantly reduce the enormous load that AI has on the power grids.
At the same time, AI models could become even more powerful. The researchers are at least convinced that optical AI calculations will soon be a central component of every AI chip. Their concept shows that the solutions for some of the biggest problems in technology could be in a radical departure of old concepts.
Also interesting:
- Why e-cars consume more energy on the highway
- Researchers develop cheap lithium-ion battery without cobalt and nickel
- Why there would probably be no AI without pigeons
- Spotify: bypass the price increase and change in basic
The contribution of light chip could drastically reduce the energy consumption of AI first appeared on basic thinking. Follow us too Google News and Flipboard Or subscribe to our update newsletter.
As a Tech Industry expert, I believe that the development of a light chip that could drastically reduce energy consumption from AI is a significant advancement in the field of technology. AI applications typically require a large amount of energy to run efficiently, which can be costly and environmentally unsustainable.
By utilizing a light chip that is specifically designed to reduce energy consumption, we can not only improve the performance of AI systems but also lower operating costs and decrease our carbon footprint. This innovation has the potential to revolutionize the way we approach AI development and deployment, making it more accessible and sustainable for a wider range of industries and applications.
I am excited to see how this technology evolves and how it can be integrated into existing AI systems to create more efficient and environmentally friendly solutions. The potential impact of a light chip on energy consumption from AI is promising, and I look forward to seeing the positive effects it can have on the tech industry as a whole.
Credits