Anshumali Shrivastava, Shabnam Daghaghi, Nicholas Meisburger
Rice University computer scientists have demonstrated artificial intelligence (AI) software that runs on commodity processors and trains deep neural networks 15 times faster than platforms based on graphics processors.
“The cost of training is the actual bottleneck in AI,” said Anshumali Shrivastava, an assistant professor of computer science at Rice’s Brown School of Engineering. “Companies are spending millions of dollars a week just to train and fine-tune their AI workloads.”
Shrivastava and collaborators from Rice and Intel will present research that addresses that bottleneck April 8 at the machine learning systems conference MLSys.
Deep neural networks (DNN) are a powerful form of artificial intelligence that can outperform humans at some tasks. DNN training is typically a series of matrix multiplication operations, an ideal workload for graphics processing units (GPUs), which cost about three times more than general purpose central processing units (CPUs).
“The whole industry is fixated on one kind of improvement — faster matrix multiplications,” Shrivastava said. “Everyone is looking at specialized hardware and architectures to push matrix multiplication. People are now even talking about having specialized hardware-software stacks for specific kinds of deep learning. Instead of taking an expensive algorithm and throwing the whole world of system optimization at it, I’m saying, ‘Let’s revisit the algorithm.’”
Shrivastava’s lab did that in 2019, recasting DNN training as a search problem that could be solved with hash tables. Their “sub-linear deep learning engine” (SLIDE) is specifically designed to run on commodity CPUs, and Shrivastava and collaborators from Intel showed it could outperform GPU-based training when they unveiled it at MLSys 2020.
The study they’ll present this week at MLSys 2021 explored whether SLIDE’s performance could be improved with vectorization and memory optimization accelerators in modern CPUs.
“Hash table-based acceleration already outperforms GPU, but CPUs are also evolving,” said study co-author Shabnam Daghaghi, a Rice graduate student. “We leveraged those innovations to take SLIDE even further, showing that if you aren’t fixated on matrix multiplications, you can leverage the power in modern CPUs and train AI models four to 15 times faster than the best specialized hardware alternative.”
Study co-author Nicholas Meisburger, a Rice undergraduate, said “CPUs are still the most prevalent hardware in computing. The benefits of making them more appealing for AI workloads cannot be understated.”
Original Article: Rice, Intel optimize AI training for commodity hardware
More from: Rice University
The Latest Updates from Bing News & Google News
Go deeper with Bing News on:
Deep neural networks
- WiMi Develops Deep Learning-Based 3D Target Recognition Algorithm System
Deep learning can simulate the human brain's visual system with its characteristics of passing from concrete to abstraction and multiple layers. By building multilayer neural networks, the system ...
- AI will soon become impossible for humans to comprehend – the story of neural networks tells us why
The mystery is even coded into the very form and discourse of the neural network. They come with deeply piled layers – hence the phrase deep learning – and within those depths are the even more ...
- Mapping neural connections in OCD patients can predict improvements after deep brain stimulation
Deep brain stimulation (DBS) is a promising therapy for treatment-resistant obsessive-compulsive disorder (OCD).
- An architecture that combines deep neural networks and vector-symbolic models
Researchers at IBM Research Zürich and ETH Zürich have recently created a new architecture that combines two of the most renowned artificial intelligence approaches, namely deep neural networks and ...
- Imaging brain connections can predict improvements in OCD patients after deep brain stimulation
Deep brain stimulation (DBS) is a promising therapy for treatment-resistant obsessive-compulsive disorder (OCD). A first-of-its-kind collaborative study led by researchers at Texas Children's Hospital ...
Go deeper with Google Headlines on:
Deep neural networks
[google_news title=”” keyword=”deep neural networks” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]
Go deeper with Bing News on:
Training deep neural networks
- AI will soon become impossible for humans to comprehend – the story of neural networks tells us why
As the layers multiplied, deep learning plumbed new depths. The neural network is trained using training data that, Hardesty explained, “is fed to the bottom layer – the input layer – and it passes ...
- A method for designing neural networks optimally suited for certain tasks
Neural networks, a type of machine-learning model, are being used to help humans complete a wide variety of tasks, from predicting if someone's credit score is high enough to qualify for a loan to ...
- An architecture that combines deep neural networks and vector-symbolic models
Researchers at IBM Research Zürich and ETH Zürich have recently created a new architecture that combines two of the most renowned artificial intelligence approaches, namely deep neural networks and ...
- ChatGPT sucks at Wordle, it won’t even help me cheat
Wordle is one of the most popular games around but with only one attempt per day, there can be pressure to get it right. And ChatGPT can't help you.
- The debate over neural network complexity: Does bigger mean better?
The ideal size and intricacy of neural networks remain a matter of debate. Does neural network complexity matter? AI experts weigh in.
Go deeper with Google Headlines on:
Training deep neural networks
[google_news title=”” keyword=”training deep neural networks” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]
