Researchers from the Institute of Industrial Science at The University of Tokyo designed and built specialized computer hardware consisting of stacks of memory modules arranged in a 3D-spiral for artificial intelligence (AI) applications. This research may open the way for the next generation of energy efficient AI devices.
Machine learning is a type of AI that allows computers to be trained by example data to make predictions for new instances. For example, a smart speaker algorithm like Alexa can learn to understand your voice commands, so it can understand you even when you ask for something for the first time.
Researchers from The University of Tokyo create a new integrated 3D-circuit architecture for AI applications with spiraling stacks of memory modules, which may help lead to specialized machine-learning hardware that uses much less electricity
Credit: Institute of Industrial Science, The University of Tokyo
However, AI tends to require a great deal of electrical energy to train, which raises concerns about adding to climate change.
Now, scientists from the Institute of Industrial Science at The University of Tokyo have developed a novel design for stacking resistive random-access memory modules with oxide semiconductor (IGZO) access transistor in a three-dimensional spiral.
Having on-chip nonvolatile memory placed close to the processors makes the machine learning training process much faster and more energy efficient. This is because electrical signals have a much shorter distance to travel compared with conventional computer hardware. Stacking multiple layers of circuits is a natural step, since training the algorithm often requires many operations to be run in parallel at the same time.
"For these applications, each layer's output is typically connected to the next layer's input. Our architecture greatly reduces the need for interconnecting wiring," says first author Jixuan Wu.
The team was able to make the device even more energy efficient by implementing a system of binarized neural networks. Instead of allowing the parameters to be any number, they are restricted to be either +1 or -1.
This both greatly simplifies the hardware used, as well as compressing the amount of data that must be stored. They tested the device using a common task in AI, interpreting a database of handwritten digits. The scientists showed that increasing the size of each circuit layer could enhance the accuracy of the algorithm, up to a maximum of around 90%.
"In order to keep energy consumption low as AI becomes increasingly integrated into daily life, we need more specialized hardware to handle these tasks efficiently," explains Senior author Masaharu Kobayashi.
This work is an important step towards the "internet of things," in which many small AI-enabled appliances communicate as part of an integrated "smart-home."
About Institute of Industrial Science (IIS), the University of Tokyo
Institute of Industrial Science (IIS), the University of Tokyo is one of the largest university-attached research institutes in Japan.
More than 120 research laboratories, each headed by a faculty member, comprise IIS, with more than 1,000 members including approximately 300 staff and 700 students actively engaged in education and research. Our activities cover almost all the areas of engineering disciplines. Since its foundation in 1949, IIS has worked to bridge the huge gaps that exist between academic disciplines and real-world applications.
Masaharu Kobayashi | EurekAlert!
New method for simulating yarn-cloth patterns to be unveiled at ACM SIGGRAPH
09.07.2020 | Association for Computing Machinery
Virtual Reality Environments for the Home Office
09.07.2020 | Universität Stuttgart
New insight into the spin behavior in an exotic state of matter puts us closer to next-generation spintronic devices
Aside from the deep understanding of the natural world that quantum physics theory offers, scientists worldwide are working tirelessly to bring forth a...
Kiel physics team observed extremely fast electronic changes in real time in a special material class
In physics, they are currently the subject of intensive research; in electronics, they could enable completely new functions. So-called topological materials...
Solar cells based on perovskite compounds could soon make electricity generation from sunlight even more efficient and cheaper. The laboratory efficiency of these perovskite solar cells already exceeds that of the well-known silicon solar cells. An international team led by Stefan Weber from the Max Planck Institute for Polymer Research (MPI-P) in Mainz has found microscopic structures in perovskite crystals that can guide the charge transport in the solar cell. Clever alignment of these "electron highways" could make perovskite solar cells even more powerful.
Solar cells convert sunlight into electricity. During this process, the electrons of the material inside the cell absorb the energy of the light....
Empa researchers have succeeded in applying aerogels to microelectronics: Aerogels based on cellulose nanofibers can effectively shield electromagnetic radiation over a wide frequency range – and they are unrivalled in terms of weight.
Electric motors and electronic devices generate electromagnetic fields that sometimes have to be shielded in order not to affect neighboring electronic...
A promising operating mode for the plasma of a future power plant has been developed at the ASDEX Upgrade fusion device at Max Planck Institute for Plasma...
07.07.2020 | Event News
02.07.2020 | Event News
19.05.2020 | Event News
10.07.2020 | Life Sciences
10.07.2020 | Materials Sciences
10.07.2020 | Life Sciences