English

Scientists demonstrate a new optical neural network training method that can crush electronic microprocessors

1223
2023-09-27 15:24:41
See translation

The current deep neural network system (such as ChatGPT) can quickly improve energy efficiency by 100 times in training, and "future improvements will greatly increase by several orders of magnitude. Scientists from MIT and other institutions have demonstrated a new optical neural network training method that can crush state-of-the-art electronic microprocessors.

Moreover, the computational density of the demonstrated system is about two orders of magnitude higher than that of Nvidia, Google, or Graphcore systems.

Basically, this means that the most advanced models can be trained with 100 times less energy and occupy less space at the same speed.

Artificial neural networks mimic the way biological brains process information. These artificial intelligence systems aim to learn, combine, and summarize information from big datasets, reshaping the field of information processing. Current applications include images, objects, speech recognition, games, medicine, and physical chemistry.

The current artificial intelligence model has reached hundreds of billions of artificial neurons, showing exponential growth and posing challenges to current hardware capabilities.

This paper demonstrates that optical neural network (ONN) methods with high clock speed, parallelism, and low loss data transmission can overcome current limitations.

Our technology opens up a path for large-scale optoelectronic processors to accelerate machine learning tasks from data centers to decentralized edge devices, "the paper wrote.

The ONN method is expected to alleviate the bottlenecks of traditional processors, such as the number of transistors, data mobility energy consumption, and semiconductor size. ONN uses light, which can carry a large amount of information simultaneously due to its wide bandwidth and low data transmission loss. In addition, many photonic circuits can be integrated to expand the system.

In order to move light for calculation, the team led by MIT utilized many laser beams, which were described as "using mass-produced micrometer scale vertical cavity surface emitting lasers for neuron coding".

The researchers explained, "Our scheme is similar to the 'axon synapse dendrite' structure in biological neurons
They believe that the demonstrated system can be expanded through mature wafer level manufacturing processes and photon integration.

Dirk Englund, Associate Professor of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology and the head of this work, explained to SciTechDaily that the size of models such as ChatGPT is limited by the capabilities of today's supercomputers. Therefore, training larger models is not economically feasible.

He claimed, "Our new technology can make it possible to cross machine learning models, otherwise it would not be possible in the near future.

This paper titled "Deep Learning Using Coherent VCSEL Neural Networks" was published by a large team of scientists. This work has received support from the Army Research Office, NTT Research, and NTT Netcast Awards, as well as financial support from the Volkswagen Foundation. The three researchers of the team have applied for patents related to this technology.

Source: Laser Network

Related Recommendations
  • Xi'an Institute of Optics and Fine Mechanics has made new progress in the field of metasurface nonlinear photonics

    Recently, the Research Group of Nonlinear Photonics Technology and Applications in the State Key Laboratory of Transient Optics and Photonics Technology of Xi'an Institute of Optics and Fine Mechanics has made important progress in the field of super surface nonlinear photonics. Relevant research results were published in the internationally famous journal Nanoscale Horizons. The first author of t...

    2024-09-27
    See translation
  • Coherent Unifies Ultrafast Laser Business at the Glasgow Center of Excellence

    Recently, Coherent, an American laser system solution provider, announced that all of the company's ultra fast laser business, including the manufacturing of all picosecond and femtosecond lasers, will be unified in one place: the Ultra Fast Center of Excellence in Glasgow, Scotland.Previously, Coherent's Ultra Fast Center of Excellence located in Glasgow was already a state-of-the-art mass produc...

    2023-09-22
    See translation
  • Google works with magic leap on AR optics and manufacturing

    In the 2010s, Magic leap is one of the most hyped augmented reality companies, with a lot of money, including from Google. When the magic leap one headset was introduced in 2018, it was not a technological breakthrough in display technology that was once derided. Since then, Magic leap has persevered and has now signed a "multifaceted strategic technology partnership" with Google.Google announced ...

    2024-05-31
    See translation
  • New machine learning algorithm accurately decodes molecular optical 'fingerprints'

    Recently, a research team from Rice University in the United States developed a new machine learning algorithm - Peak Sensitive Elastic Network Logistic Regression (PSE-LR). This algorithm is adept at interpreting the unique optical characteristics of molecules, materials, and disease biomarkers, which can help achieve faster and more accurate medical diagnosis and sample analysis. The relevant pa...

    05-09
    See translation
  • Brother Australia launches innovative professional monochrome laser series

    Brother Australia is a renowned printing manufacturer that has expanded its product portfolio by launching its latest innovative commercial machine series, the professional monochrome laser machine series. These extraordinary devices are designed to extend the lifespan of commercial printing cycles and improve productivity. Due to their sturdy components and durable consumables, these extraordinar...

    2024-03-21
    See translation