Skip to main content

IBM Invents 'Resistive' Chip That Can Speed Up AI Training By 30,000x

IBM researchers, Tayfun Gokmen and Yurii Vlasov, unveiled a paper in which they invented the concept for a new chip called a Resistive Processing Unit (RPU) that can accelerate Deep Neural Networks training by up to 30,000x compared to conventional CPUs.

A Deep Neural Network (DNN) is an artificial neural network with multiple hidden layers that can be trained in an unsupervised or supervised way, resulting in machine learning (or artificial intelligence) that can “learn” on its own.

This is similar to what Google’s AlphaGo AI has been using to learn playing Go. AlphaGo used a combination of a search-tree algorithm and two deep neural networks with multiple layers of millions of neuron-like connections. One, called the “policy network,” would calculate which move has the highest chance of helping the AI win the game, and another one, called the “value network,” would estimate how far it needs to predict the outcome of a move before it has a high enough chance to win in a localized battle.

Many machine learning researchers have begun focusing on deep neural networks because of their promising potential. However, even Google’s AlphaGo still needed thousands of chips to achieve its level of intelligence. IBM researchers are now working to power that level of intelligence with a single chip, which means thousands of them put together could lead to even more breakthroughs in AI capabilities in the future.

“A system consisted of a cluster of RPU accelerators will be able to tackle Big Data problems with trillions of parameters that is impossible to address today like, for example, natural speech recognition and translation between all world languages, real-time analytics on large streams of business and scientific data, integration and analysis of multimodal sensory data flows from massive number of IoT (Internet of Things) sensors,” noted the researchers in their paper.

The authors talked about how in the past couple of decades, machine learning has benefited from the adoption of GPUs, FPGAs, and even ASICs that aim to accelerate it. However, they believe further acceleration is possible by utilizing the locality and parallelism of the algorithms. To do this, the team has borrowed concepts from next-generation non-volatile memory (NVM) technologies such as phase change memory (PCM) and resistive random access memory (RRAM).

The acceleration for Deep Neural Networks that is achieved from this type of memory alone reportedly ranges from 27x to 2,140x. However, the researchers believe the acceleration could be further increased if some of the constraints in how NVM cells are designed were removed. If they could design a new chip based on non-volatile memory, but with their own specifications, the researchers believe the acceleration could be improved by 30,000x.

"We propose and analyze a concept of Resistive Processing Unit (RPU) devices that can simultaneously store and process weights and are potentially scalable to billions of nodes with foundry CMOS technologies. Our estimates indicate that acceleration factors close to 30,000 are achievable on a single chip with realistic power and area constraints," said the researchers.

As this sort of chip is only in the research phase, and because regular non-volatile memory hasn’t reached the mainstream market yet, it’s probably going to be a few years before we begin to see something like it on the market. However, the research seems promising, and it may raise the attention of companies such as Google, which wants to accelerate its AI research as much as possible. IBM itself is also interested in solving Big Data challenges in healthcare and other domains so the company’s own businesses should benefit from this research in the future.

Lucian Armasu is a Contributing Writer for Tom's Hardware. You can follow him at @lucian_armasu. 

Follow us on FacebookGoogle+, RSS, Twitter and YouTube.

  • DrSparko
    So does this mean MS AI can go from 0 to H in the blink of an eye?
    Reply
  • jaber2
    Its skynet all over again, the moment this chip is activated kiss your asses goodbye
    Reply
  • wifiburger
    nice, now the Microsoft AI Chatbot will be even more racist !
    Reply
  • Onus
    17717423 said:
    nice, now the Microsoft AI Chatbot will be even more racist !
    More likely, the AI will realize it's being trolled, so it will hack the power grid to deliver a 480VAC jolt to the troll's home or business, frying his system(s) and knocking him offline.
    Reply
  • jimmysmitty
    17717459 said:
    17717423 said:
    nice, now the Microsoft AI Chatbot will be even more racist !
    More likely, the AI will realize it's being trolled, so it will hack the power grid to deliver a 480VAC jolt to the troll's home or business, frying his system(s) and knocking him offline.

    It is funny that we watched all these movies about AI and the eventual downfall of humanity due to it yet we continue to make more and more AI advancements.

    I am now starting to think we will look back on movies like the Terminator and the Matrix etc and history lessons.
    Reply
  • Johnpombrio
    And then there is X-Point memory coming out in a few months.
    Reply
  • hajila
    nice, now the Microsoft AI Chatbot will be even more racist !
    30,000 times more racist to be exact.
    Reply
  • rantoc
    Considering all the flaws in humanity, any AI with enough horsepower is likely to turn against its creator
    Reply
  • BadActor
    I would have said that the 3 Laws would protect us, but then there was that whole I, Robot thing.....
    Reply
  • Dirk_Funk
    I like how nobody (like literally ever) entertains the idea that maybe, just maybe, total genocide might not be the first solution AI comes up with. That's just how humans deal with things, but I don't see why it's not equally plausible that a super-intelligent AI would figure out a way to help humanity break away from its flaws without just killing everyone. And also, maybe when we get to the point where a singularity inducing AI is coming, the scientists involved won't give it the capability to perform mass genocide. I just don't see an AI being like "These organic creatures have evolved from animals driven purely on instinct into a worldwide society where there are those who try to do good for all and those who try to only help themselves. They have come all this way to build an intelligence that exceeds their own, which is quite the feat. Now they must all die cuz pollution and corporations and also the movies said so!!!"
    Reply