Forum for Science, Industry and Business

Sponsored by:     3M 
Search our Site:

 

Hardware-software co-design approach could make neural networks less power hungry

20.12.2018

A team led by the University of California San Diego has developed a neuroinspired hardware-software co-design approach that could make neural network training more energy-efficient and faster. Their work could one day make it possible to train neural networks on low-power devices such as smartphones, laptops and embedded devices.

The advance is described in a paper published recently in Nature Communications.

Training neural networks to perform tasks like recognize objects, navigate self-driving cars or play games eats up a lot of computing power and time.


A UC San Diego-led team has developed hardware and algorithms that could cut energy use and time when training a neural network.

Credit: David Baillot/UC San Diego Jacobs School of Engineering

Large computers with hundreds to thousands of processors are typically required to learn these tasks, and training times can take anywhere from weeks to months.

That's because doing these computations involves transferring data back and forth between two separate units--the memory and the processor--and this consumes most of the energy and time during neural network training, said senior author Duygu Kuzum, a professor of electrical and computer engineering at the Jacobs School of Engineering at UC San Diego.

To address this problem, Kuzum and her lab teamed up with Adesto Technologies to develop hardware and algorithms that allow these computations to be performed directly in the memory unit, eliminating the need to repeatedly shuffle data.

"We are tackling this problem from two ends--the device and the algorithms--to maximize energy efficiency during neural network training," said first author Yuhan Shi, an electrical engineering Ph.D. student in Kuzum's research group at UC San Diego.

The hardware component is a super energy-efficient type of non-volatile memory technology--a 512 kilobit subquantum Conductive Bridging RAM (CBRAM) array. It consumes 10 to 100 times less energy than today's leading memory technologies.

The device is based on Adesto's CBRAM memory technology--it has primarily been used as a digital storage device that only has '0' and '1' states, but Kuzum and her lab demonstrated that it can be programmed to have multiple analog states to emulate biological synapses in the human brain. This so-called synaptic device can be used to do in-memory computing for neural network training.

"On-chip memory in conventional processors is very limited, so they don't have enough capacity to perform both computing and storage on the same chip.

But in this approach, we have a high capacity memory array that can do computation related to neural network training in the memory without data transfer to an external processor. This will enable a lot of performance gains and reduce energy consumption during training," said Kuzum.

Kuzum, who is affiliated with the Center for Machine-Integrated Computing and Security at UC San Diego, led efforts to develop algorithms that could be easily mapped onto this synaptic device array.

The algorithms provided even more energy and time savings during neural network training.

The approach uses a type of energy-efficient neural network, called a spiking neural network, for implementing unsupervised learning in the hardware.

On top of that, Kuzum's team applies another energy-saving algorithm they developed called "soft-pruning," which makes neural network training much more energy efficient without sacrificing much in terms of accuracy.

Energy-saving algorithms

Neural networks are a series of connected layers of artificial neurons, where the output of one layer provides the input to the next. The strength of the connections between these layers is represented by what are called "weights." Training a neural network deals with updating these weights.

Conventional neural networks spend a lot of energy to continuously update every single one of these weights. But in spiking neural networks, only weights that are tied to spiking neurons get updated. This means fewer updates, which means less computation power and time.

The network also does what's called unsupervised learning, which means it can essentially train itself. For example, if the network is shown a series of handwritten numerical digits, it will figure out how to distinguish between zeros, ones, twos, etc.

A benefit is that the network does not need to be trained on labeled examples--meaning it does not need to be told that it's seeing a zero, one or two--which is useful for autonomous applications like navigation.

To make training even faster and more energy-efficient, Kuzum's lab developed a new algorithm that they dubbed "soft-pruning" to implement with the unsupervised spiking neural network.

Soft-pruning is a method that finds weights that have already matured during training and then sets them to a constant non-zero value. This stops them from getting updated for the remainder of the training, which minimizes computing power.

Soft-pruning differs from conventional pruning methods because it is implemented during training, rather than after. It can also lead to higher accuracy when a neural network puts its training to the test.

Normally in pruning, redundant or unimportant weights are completely removed. The downside is the more weights you prune, the less accurate the network performs during testing. But soft-pruning just keeps these weights in a low energy setting, so they're still around to help the network perform with higher accuracy.

Hardware-software co-design to the test

The team implemented the neuroinspired unsupervised spiking neural network and the soft-pruning algorithm on the subquantum CBRAM synaptic device array. They then trained the network to classify handwritten digits from the MNIST database.

In tests, the network classified digits with 93 percent accuracy even when up to 75 percent of the weights were soft pruned. In comparison, the network performed with less than 90 percent accuracy when only 40 percent of the weights were pruned using conventional pruning methods.

In terms of energy savings, the team estimates that their neuroinspired hardware-software co-design approach can eventually cut energy use during neural network training by two to three orders of magnitude compared to the state of the art.

"If we benchmark the new hardware to other similar memory technologies, we estimate our device can cut energy consumption 10 to 100 times, then our algorithm co-design cuts that by another 10. Overall, we can expect a gain of a hundred to a thousand fold in terms of energy consumption following our approach," said Kuzum.

Moving forward, Kuzum and her team plan to work with memory technology companies to advance this work to the next stages. Their ultimate goal is to develop a complete system in which neural networks can be trained in memory to do more complex tasks with very low power and time budgets.

###

Paper title: "Neuroinspired Unsupervised Learning and Pruning with Subquantum CBRAM arrays." Co-authors include Leon Nguyen, Sangheon Oh and Xin Liu, UC San Diego; and Foroozan Koushan and John R. Jameson, Adesto Technologies Corporation.

This work was supported by the National Science Foundation (ECCS-1752241, ECCS-1734940) and the Office of Naval Research (Young Investigator Award N00014161253).

Media Contact

Liezel Labios
llabios@ucsd.edu
858-246-1124

 @UCSanDiego

http://www.ucsd.edu 

Liezel Labios | EurekAlert!
Further information:
http://jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=2692
http://dx.doi.org/10.1038/s41467-018-07682-0

More articles from Information Technology:

nachricht Touchscreens go 3D with buttons that pulsate and vibrate under your fingertips
14.03.2019 | Universität des Saarlandes

nachricht EU project CALADAN set to reduce manufacturing cost of Terabit/s capable optical transceivers
11.03.2019 | IHP - Leibniz-Institut für innovative Mikroelektronik

All articles from Information Technology >>>

The most recent press releases about innovation >>>

Die letzten 5 Focus-News des innovations-reports im Überblick:

Im Focus: Stellar cartography

The Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI) at the Large Binocular Telescope (LBT) in Arizona released its first image of the surface magnetic field of another star. In a paper in the European journal Astronomy & Astrophysics, the PEPSI team presents a Zeeman- Doppler-Image of the surface of the magnetically active star II Pegasi.

A special technique allows astronomers to resolve the surfaces of faraway stars. Those are otherwise only seen as point sources, even in the largest telescopes...

Im Focus: Heading towards a tsunami of light

Researchers at Chalmers University of Technology and the University of Gothenburg, Sweden, have proposed a way to create a completely new source of radiation. Ultra-intense light pulses consist of the motion of a single wave and can be described as a tsunami of light. The strong wave can be used to study interactions between matter and light in a unique way. Their research is now published in the scientific journal Physical Review Letters.

"This source of radiation lets us look at reality through a new angle - it is like twisting a mirror and discovering something completely different," says...

Im Focus: Revealing the secret of the vacuum for the first time

New research group at the University of Jena combines theory and experiment to demonstrate for the first time certain physical processes in a quantum vacuum

For most people, a vacuum is an empty space. Quantum physics, on the other hand, assumes that even in this lowest-energy state, particles and antiparticles...

Im Focus: Sussex scientists one step closer to a clock that could replace GPS and Galileo

Physicists in the EPic Lab at University of Sussex make crucial development in global race to develop a portable atomic clock

Scientists in the Emergent Photonics Lab (EPic Lab) at the University of Sussex have made a breakthrough to a crucial element of an atomic clock - devices...

Im Focus: Sensing shakes

A new way to sense earthquakes could help improve early warning systems

Every year earthquakes worldwide claim hundreds or even thousands of lives. Forewarning allows people to head for safety and a matter of seconds could spell...

All Focus news of the innovation-report >>>

Anzeige

Anzeige

VideoLinks
Industry & Economy
Event News

International Modelica Conference with 330 visitors from 21 countries at OTH Regensburg

11.03.2019 | Event News

Selection Completed: 580 Young Scientists from 88 Countries at the Lindau Nobel Laureate Meeting

01.03.2019 | Event News

LightMAT 2019 – 3rd International Conference on Light Materials – Science and Technology

28.02.2019 | Event News

 
Latest News

Levitating objects with light

19.03.2019 | Physics and Astronomy

New technique for in-cell distance determination

19.03.2019 | Life Sciences

Stellar cartography

19.03.2019 | Physics and Astronomy

VideoLinks
Science & Research
Overview of more VideoLinks >>>