Artificial intelligence is complex. We know complexity.
Artificial intelligence (AI) and machine learning are technologies that mimic the capabilities of the human brain, which has over 100 trillion neural connections. That’s an enormously complex task.
For over 40 years, Cray has been solving some of the most complex computing problems in science and industry. We’re the natural choice to supply the high-powered computing systems you need to make artificial intelligence, machine learning and deep learning work for you.
We help you envision the potential of artificial intelligence.
Cars that drive themselves. Phones that recognize human faces. Software you can chat with. AI and deep learning have captured the public imagination in consumer-friendly applications like these.
The techniques that make these applications possible have been honed behind the scenes in less visible but no less impressive technologies such as medical imaging, cybersecurity, climate modeling, seismic processing, computational fluid dynamics and quantitative finance.
Cray has built the high-performance computing systems that have driven success in these and other complex applications. We understand the challenges of artificial intelligence in all kinds of uses.
We give you the power to explore.
Cray is a partner that can help you start and realize your first success with AI. It’s a journey, and we’re on it with you.
Big Data Maturity
Analytics allow you to sense and comprehend what is happening; machine learning and deep learning allow you to predict and act on that prediction with confidence. Underpinning both is the presence of big data that is captured, managed and optimized. The Cray® Urika®-GX agile analytics platform is an integrated system designed for complex machine workflows:
- Starting with start with streaming data collection using tools like Apache Kafka®
- Advancing to data cleansing, preparation and enrichment using data Apache Spark™
- Performing machine learning using tools like MLLib
- Advanced pattern matching and discovery using the Cray Graph Engine
Machine Learning Exploration
Machine learning and deep learning are everywhere, but where do you start? To help you accelerate your progression along an AI journey, we’ve created the Cray Accel AI initiative to help organizations start, learn, experience and succeed with machine and deep learning.
Cray Accel AI Fast Start Offerings
Cray Accel AI fast-start configurations range from a starter system ideal for AI exploration to a complete, production-level Cray cluster supercomputer for training and inference. Components include the Cray CS-Storm accelerated GPU system featuring NVIDIA® Tesla® V100 GPU accelerators powered by the NVIDIA® Volta™ GPU architecture, and a comprehensive deep learning and analytics software environment from Bright Computing.
Cray Accel AI Lab
The Cray Accel AI Lab serves as a focal point for 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 AI experts from Cray’s engineering teams, Cray partners and customers can learn and experience the latest advanced optimization techniques for deep learning frameworks.
Bringing HPC Best Practices to AI
Machine learning and, more specifically, deep learning are, at their core, high-performance computing problems — problems that Cray has a 40-plus-year history of solving. It why we bring HPC best practices to AI.
Cray® XC™ Series Supercomputers and the Urika-XC Software Suite
The Cray Urika-XC analytics and AI software suite, which brings graph analytics, deep learning, and robust big data analytics tools to our flagship line of Cray XC supercomputers, now includes the TensorFlow computational framework and enhancements to the Cray software environment that are specifically designed to accelerate machine learning frameworks. The new Cray Programming Environment (CPE) Machine Learning Library delivers unprecedented scaling by leveraging the Aries™ interconnect, and it eliminates many of the mundane administrative tasks required for distributed deep learning. With this enhanced performance, data scientists can train neural networks at scale on a Cray XC supercomputer — leveraging either CPUs or GPUs — to more than 500 nodes.