Urika-XC

Eliminate AI and Analytics Complexity for Faster Results

Contact Us

AI and Analytics at the Speed of Supercomputing

Cray's Urika®-XC AI and big data analytics tools bring you advanced analytics, machine learning and graph tools, engineered for Cray, a Hewlett Packard Enterprise company, XC series supercomputers. Do It All

Gain new insights and solve the toughest problems across all your data on one machine — without additional hardware and without moving data from machine to machine. The Cray Urika-XC suite allows you to run big data analytics tools, machine learning and deep learning applications on the same system at the same time, on a system also running high-performance simulations and using familiar system tools and schedulers. Cray Urika-XC Analytics Software With supercomputer speed and efficiency, you can run more types of analytics and artificial intelligence workloads faster, yielding unprecedented breadth of results and maximizing your opportunity for transformative discoveries.

Get Modern

Use modern machine learning frameworks, big data analytics tools, and data science languages to power richer insights and reduce the need for scarce talent. Cray’s Urika-XC suite includes a set of powerful big data analytics, data science and AI tools:

  • Big data analytics toolswith Apache® Spark™ and the Spark ecosystem running at supercomputer scale and pbdR (Programming with Big Data using R)
  • Data science tools: Python-based tools and libraries for data scientist productivity (Anaconda and Distributed Dask) and Jupyter notebooks for end-user productivity.
  • Machine and Deep learning frameworks (TensorFlow™, PyTorch and Keras) and libraries (hyperparameter optimization and distributed training) enhanced to run at supercomputer scale (using 1,000 CPUs or GPUs) with simplified setup and administration.
  • Graph databases using Cray’s powerful Cray Graph Engine (CGE) for big data graph pattern recognition and discovery applications. For the most demanding graph problems where size and performance matter, CGE has been tested against a 400-billion-tuple collection and shows a nearly 100x performance speed-up over Spark-based GraphX.

Innovate with Deep Learning Applications at Supercomputer Scale

You can capitalize on the XC series supercomputer’s scale and performance for your deep learning research with the Cray-developed library for distributed training (Cray PE ML plugin) and hyperparameter optimization (HPO). The Cray PE ML plugin is delivered in conjunction with popular machine intelligence frameworks (TensorFlow™, PyTorch and Keras) and leverages the Aries™ interconnect to give you scaling to more than 500 nodes on both CPUs and GPUs. Cray’s HPO libraries eliminate much of the guesswork required by a data scientist to set up a complex model and reduces the time required to achieve model accuracy.

Improve TCO

With Urika-XC advanced analytics, you can make the most of your investment in Cray XC series supercomputers and reduce the need to maintain separate systems. Meet the increased demand for advanced analytics without capital investment or retraining.

Featured Resources

Cray Urika-XC Analytics & AI Software Suite

The Urika-XC software suite brings big data analytics, artificial intelligence (AI) and large-scale graph processing to the massively scalable Cray XC series supercomputer.

For Data Scientists: Cray Urika-XC Advanced Analytics Suite

If you're a data scientist, the Cray Urika-XC analytics suite offers you the complete set of tools that you're used to, many of them open source, operating at a power and scale that you've never seen before.

See All Resources

Discovery Unleashed

Cray Urika-XC analytics helps Cray XC users expand into high-performance data analytics, graph processing and artificial intelligence for data science and research.

Datacenter ManagerSupercomputer center managers need to keep up with the rapid expansion of analytics and AI and provide an infrastructure that “does it all.”

  • With the ability to run analytics, graph and AI simultaneously on the same system, supercomputer lab managers can make the most of their investments in Cray XC supercomputers and reduce the need to maintain separate systems.
  • You'll get more services for more users, and a better return on your investment.

Data scientists gain new insights and solve the toughest problems, crossing analytics and research domains, with data on one machine — without additional hardware and without moving data from machine to machine.

  • With supercomputer speed and efficiency you can run more types of analytics faster, yielding unprecedented breadth of results and maximizing your opportunity for transformative discoveries.

Urika-XC Technology

Bringing analytics, AI and graph to Cray XC series supercomputers

The Urika®-XC software suite is a set of powerful big data analytics and AI applications and tools optimized to run on the Cray XC line of supercomputing systems. They’ve been reengineered to take advantage of the unique scalable architecture of the Cray XC series, including Cray’s Aries™ network interconnect. Urika-XC software tools are integrated with the high-performance Cray Linux® Environment and managed through existing workload managers. With the Urika-XC package, analytics and AI applications benefit from the supercomputer power and scale of Cray XC systems.

Unmatched Performance and Scale for Graph Applications

Based on the graph technology in our Urika®-GX agile analytics platform, Urika-XC graph technology breaks new ground in scaling and performance for classic graph-based pattern matching and discovery applications. Our implementation of graph is an RDF triplestore database capable of supporting massive graph instances, taking advantage of the performance and scale found in the Cray XC series, and implementing W3C’s SPARQL query language for RDF, along with unique-to-Cray built-in graph functions that extend SPARQL.

Open-Source Analytics at Supercomputer Scale

Urika-XC includes the latest in open source technologies for big data analytics and AI. Integrated with standard HPC workload managers, the Urika-XC software suite brings the power of Apache® Spark™ clusters, distributed Dask clusters and pbdR to our XC series.

  • Dynamically provisions Apache Spark clusters for in-memory analytics with support for libraries (MLlib, GraphX and SparkSQL and BigDL) and languages/tools (Python, Scala and Java, sbt and Maven).
  • Supports researcher productivity with the right data science tools (Anaconda® Distribution, (CONDA® and Data Science Libraries), Python, distributed Dask)
  • Includes scaled parallel infrastructure for big data analytics using the R language (Programming with Big Data in R)
Cray Urika-XC Software Suite

Distributed Deep Learning with BigDL

You can capitalize on the XC series supercomputer’s scale and performance for your deep learning workloads with the Cray Distributed Training Framework, including unique Cray-developed distributed training and hyperparameter optimization (HPO) libraries. Delivered in conjunction with the industry-leading machine intelligence frameworks (TensorFlow™, PyTorch and Keras), the Cray Distributed Training Framework leverages the Aries™ interconnect to give you scaling to more than 500 nodes on both CPUs and GPUs.