Deep Learning

CRAY SOLUTIONS FOR DEEP LEARNING WORKFLOWS

Contact Us

Take Artificial Intelligence Deep

Deep learning takes artificial intelligence deep into the workings of the human mind.

Start with the right partner.

We've been helping visionaries like you solve the most complex computing problems for over four decades.

Using machine learning tools and techniques,
deep learning aims to solve problems the way a human could.

Just consider the complexity. The human brain has over 100 trillion neural connections. Training artificial neural networks — networks that think and learn like the human brain — takes an endless infusion of data. Data that must be prepared, trained, tested and deployed over and over again. But deep learning’s potential is as boundless as the mind. Its possible applications extend from healthcare to manufacturing to finance and more.



LEARN ALONGSIDE US

We partner with industry leaders like NVIDIA and Intel to provide you with the training to master and implement machine and deep learning. Our Cray Accel Artificial Intelligence Lab focuses on the application, incubation and advancement of supercomputing technologies in AI. With access to Cray systems powered by Intel® Xeon® Scalable processors and NVIDIA® Tesla® GPU accelerators, as well as Cray AI experts, you can learn about the latest advanced techniques for machine and deep learning frameworks.

Get started with Cray Accel AI

Deep Learning Goes from Possible to Practical

Deep learning remains nothing but a great idea without three elements: next-generation algorithms, massive amounts of data, and computing power that’s both next generation and massive. Scientists have developed the algorithms, technology has found ways to capture the data, and Cray technology transforms the possibilities into practice. Cray has been leading the way with supercomputers and software tools in simulation, modeling and data analytics for decades. With Cray technologies, you can start, speed and scale any deep learning initiative.



CRAY TECHNOLOGIES
FOR DEEP LEARNING

Workflow and scale are critical considerations for deep learning projects. Data preparation, model development and model implementation form the cornerstone of any deep learning effort. And scale underpins a data scientist’s ability to get work done quickly.

Our Accel AI™ reference configurations for deep learning, Cray® Urika® AI and Analytics suites, CS-Storm™ GPU-accelerated systems and CS500™ systems are designed for IT and data science organizational success with deep learning.

The Urika®-XC and Urika®-CS AI and Analytics suites remove complexity and underpin performance. With these suites, we have applied our supercomputing expertise to the challenge of making deep learning easier to use.

Both suites feature popular frameworks, languages and tools including Apache Spark™, Distributed Dask, Python and Jupyter Notebooks.

At its core, deep learning neural network processing — training and inference — is a high-performance computing application.

We’ve optimized the Urika suites for operation on multi-node systems for training and implementation. For model development we’ve included our own distributed hyperparameter optimization (HPO) library to help data scientists get to the right answer faster. We then paired that with best-of-breed distributed training libraries (the Cray PE ML plugin and Horovod) to make deep learning easier, leveraging the underlying power of a parallel processing system.

WHY CHOOSE CRAY FOR DEEP LEARNING

Cray has been solving the most complex computing problems in science and industry for over 45 years. Organizations big and small trust Cray experts. We’ll work closely with you to get you on the right path for your deep learning goals and then we’ll stay with you to make sure you have what you need to succeed.

See what others have achieved with Cray

LEARN FROM THE AI EXPERTS

Join us for hands-on training in AI and accelerated computing.



Cray instructors are certified to deliver NVIDIA’s Deep Learning Institute (DLI) content and lead workshops around the world. NVIDIA DLI provides hands-on training in artificial intelligence and accelerated computing to solve real-world problems. To find out more, fill out the form below

FURTHER YOUR AI KNOWLEDGE



Analytics and AI at the Speed of Supercomputing

Whether you’re just starting out or deep into AI, Cray HPC solutions help you achieve your goals.


Artificial Intelligence >

Cray Machine Learning Technology

With Cray technologies, machine learning is as boundless as the intelligence it aims to mimic.


Machine Learning >

Cray Accel Artificial Intelligence Lab

Get the training you need to master and implement machine and deep learning through the Accel AI Lab.


Accel AI Lab >

By using this site, you agree to Cray's terms of service and privacy policy.