日本語

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

931
2023-09-27 15:24:41
翻訳を見る

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

関連のおすすめ
  • Breakthrough development of terahertz quantum cascade lasers

    With the development of groundbreaking components for terahertz quantum cascade lasers, a huge leap has been made in the field of laser technology. A group of researchers have successfully designed a broadband single-chip external coupler with the potential to redefine the functionality of terahertz QCL.The new external coupler is fundamentally based on planar bimetallic waveguides. Its design is ...

    2024-01-04
    翻訳を見る
  • Tower and Fortsense have announced the launch of their highly advanced 3D imager for LiDAR

    Recently, Gaota Semiconductor announced the successful development of an advanced 3D imager based on dToF technology for LiDAR applications. The newly developed product FL6031 is based on Tower's 65nm Stacked BSI CIS platform and has pixel level hybrid bonding function. It is the first in a series of products aimed at meeting the needs of numerous deep sensing applications in the automotive, consu...

    2023-09-14
    翻訳を見る
  • HENGTONG listed on the Fortune Global 500 list of brands

    Recently, the 2024 (21st) World Brand 500 ranking list exclusively compiled by World Brand Lab was released in New York, USA. HENGTONG brand participated in the selection for the first time, standing out from more than 8000 participating brands in 32 countries worldwide and ranking 395th on the "Top 500 World Brands" list. This year, there are a total of 21 new brands on the global list, of whic...

    2024-12-17
    翻訳を見る
  • Germany and the United States jointly build a $150 million laser equipment laboratory for studying inertial fusion energy and high energy density physics

    German laser Fusion developer Marvel Fusion said it will partner with Colorado State University (CSU) on a new $150 million laser equipment lab to study inertial fusion energy and high energy density physics."It will be home to one of the most powerful laser facilities in the world and an international center for laser fusion energy and high energy density physics research," the company said in a ...

    2023-08-10
    翻訳を見る
  • Scientists decipher the code for extending the lifespan of perovskite solar technology

    The latest research led by the University of Surrey shows that alumina (Al2O3) nanoparticles can significantly enhance the lifespan and stability of perovskite solar cells, extending the service life of such high-efficiency energy devices tenfold.Although perovskite solar cells have advantages such as low cost and light weight compared to traditional silicon-based technologies, their commercial po...

    03-03
    翻訳を見る