Researchers from the University of Washington have developed new optical computing hardware for AI and machine learning that is faster and much more energy efficient than conventional electronics.
What is AI currently used for?
Artificial intelligence (AI) and machine learning are currently impacting our lives in various, small and impactful ways. For example, AI and machine learning applications recommend entertainment that we might enjoy through streaming services such as Netflix and Spotify.
It is predicted that these technologies will have an even larger impact on society in the future, through activities such as driving fully autonomous vehicles, enabling complex scientific research, and facilitating medical discoveries.
What are the drawbacks for current AI technologies?
The computers used for AI and machine learning utilise copious amounts of energy. Currently, the power needed for these technologies is doubling roughly every three to four months.
Additionally, cloud computing data centres employed by AI and machine learning applications worldwide are already devouring more electrical power per year than some small countries. Thus, this level of energy consumption is unsustainable.
How could AI technologies be made more energy and noise efficient?
The research team demonstrates an optical computing system for AI and machine learning, that not only mitigates this noise, but actually utilises some of it as an input to help enhance the creative output of the artificial neural network within the system.
“We’ve built an optical computer that is faster than a conventional digital computer,” explained Changming Wu, lead author and UW doctoral student in electrical and computer engineering. “And also, this optical computer can create new things based on random inputs generated from the optical noise that most researchers tried to evade.”
Optical computing noise effectively comes from stray light particles, or photons, that originate from the operation of lasers within the device and background thermal radiation. To target noise, scientists connected their optical computing core to a special type of machine learning network, called a Generative Adversarial Network (GAN).
How exactly was GAN developed?
The team tested several noise mitigation techniques, which included using some of the noise generated by the optical computing core to serve as random inputs for GAN.
The researchers then assigned GAN with the task of learning how to handwrite the number ‘7’ like a human would. It was found that the optical computer could not simply print out the number according to a prescribed font; it had to learn the task much like a child would, by looking at visual samples of handwriting and practicing until it could write the number correctly.
However, scientists noted that the optical computer did not have a human hand for writing, so its form of ‘handwriting’ was to generate digital images that had a style similar to the samples it had studied but were not identical to them.
Mo Li, senior author, and a UW professor of electrical and computer engineering commented: “Instead of training the network to read handwritten numbers, we trained the network to learn to write numbers, mimicking visual samples of handwriting that it was trained on.
“We, with the help of our computer science collaborators at Duke University, also showed that the GAN can mitigate the negative impact of the optical computing hardware noises by using a training algorithm that is robust to errors and noises. More than that, the network actually uses the noises as random input that is needed to generate output instances.”
After learning from handwritten samples of the number seven, which were from a standard AI-training image set, the GAN practiced writing ‘7’ until it could do it successfully. Along the way, it developed its own distinct writing style and could write numbers from one to 10 in computer simulations.
How is GAN going to be developed for future applications?
Researchers intend to build this device at a larger scale using the current semiconductor manufacturing technology available. This means that instead of constructing the next version of the device in a lab, the team plans to use an industrial semiconductor foundry to achieve wafer-scale technology.
Scientists hypothesise that building a larger-scale device will further improve performance and allow the research team to do more complex tasks beyond handwriting generation, such as creating artwork and even videos.
Li concluded: “This optical system represents a computer hardware architecture that can enhance the creativity of artificial neural networks used in AI and machine learning, but more importantly, it demonstrates the viability for this system at a large scale where noise and errors can be mitigated and even harnessed.
“AI applications are growing so fast that in the future, their energy consumption will be unsustainable. This technology has the potential to help reduce that energy consumption, making AI and machine learning environmentally sustainable — and very fast, achieving higher performance overall.”
