Software for Cray XC Series Supercomputers
Driving Maximum Program Performance, System Uptime and Big Data Analytics
Programming Productivity and Performance – We take a comprehensive approach to our XC series software programming environment, delivering a package of HPC-optimized programming tools, Linux® OS and a massively scalable analytics suite. We also work with domain experts to bring you innovative partner solutions.
Maximizing Uptime with System Management – To get the most workload production out of your supercomputing system, Cray provides a robust system management stack with a broad variety of utilities and support, optimizing production HPC environment reliability.
Analytics at the Speed of Supercomputing – With Cray Urika-XC analytics, XC users can unlock sophisticated big data analytics at the speed of supercomputing using modern tools and languages. With no need to purchase purpose-built analytics solutions, the Cray Urika-XC software suite enables customers to get more value from their investment in Cray XC systems.
Comprehensive Programming Environment Support for:
- Programming languages
- Programming models
- I/O libraries
- Optimized scientific libraries
- Partner middleware
- Partner ISV codes and
Cray System Management Software Drives:
- More system uptime
- More workload productivity and throughput
- More efficient upgrades and updates while in operation
- More rollback protection, less lost operation or downtime
- More ease-of-use leveraging industry standardization
Urika-XC Analytics Tools Help You:
- Unlock sophisticated big data analytics
- Maximize your investment — no need for a separate analytics machine
- Benefit from a robust set of analytics tools including Apache® Spark™, Cray Graph Engine, Python with distributed Dask, BigDL Distributed Deep Learning and many more
- Integrate your hardware and software
Cray Linux® Environment
Cray XC series systems utilize the Cray Linux Environment (CLE). CLE is a suite of high performance software that includes a Linux-based operating system designed to run large, complex applications and scale efficiently to more than 500,000 processor cores. The Linux environment features a compute kernel that can be configured to match different workloads. When running highly scalable custom applications, the compute nodes can be run in a lightweight mode, ensuring that operating system services do not interfere with application scalability. This design ensures virtually nothing stands between your scalable application and your hardware.
With CLE many independent software vendor (ISV) applications are available "out of the box," benefiting from Cray's network and MPI libraries through MPICH application binary interface compatibility. In addition, selected ISV applications are highly tuned and available under Extreme Scalability Mode (ESM).
Examples of Typical Applications Available Under ESM
A few of the HPC industry applications that may be run on Cray XC systems include:
- Astrophysics: Cactus, Chimera, Gadget, QCD-Milc, ZEUS-MP
- Chemistry/Materials: ABINIT, AMBER, CPMD, DCA++, GAMESS, Gromacs, LAMMPS, LS3DF, LSMS, Madness, NWCHEM, PARATEC, Qbox, Quantum Espresso, Siesta, TBMD, VASP
- Climate/Weather: 4DVAR, CAM, CCSM, HIRLAM, HYCOM, MM5, POP, WRF
- Fluid Mechanics/Combustion: AcuSolve, DNS, OpenFOAM, Overflow, PFLOTRAN, S3D
- Fusion: Aorsa, GTC
- Solid Mechanics: CTH, LS-DYNA
The CLE programming environment includes tools designed to complement and enhance each other, resulting in a rich, easy-to-use programming environment that facilitates the development of scalable applications.
- Parallel programming using OpenMP and OpenACC within the node, and MPI, Cray SHMEM, UPC and Fortran across the system
- MPI standard, optimized to take advantage of the scalable interconnect in Cray XC series systems
- Various MPI libraries supported under Cluster Compatibility Mode
- Support for ISO standard languages including: Fortran 2008 (including parallel programming using coarrays), C 11, C++ 14
- Support for shared memory programming using Unified Parallel C
- High performance optimized math libraries of BLAS, FFTs, LAPACK, ScaLAPACK, SuperLU and Cray scientific libraries
- Cray Performance Tool Suite with CrayPAT™, Cray Apprentice2™ and Cray Reveal™
Shifter (Docker) Container Support
To provide the flexibility and performance of Docker containers, Cray XC series systems include support for Shifter — a Docker-like container environment designed for the demands of a massive, multi-processor supercomputer. With Shifter, Cray brings the application portability and ease-of-use of containers for highly-scalable applications to Cray’s advanced supercomputers.
With Shifter as part of Cray’s official software distribution, XC series users can implement software containers that package an entire software stack — application codes, shared libraries, base operating system files, user environment variables, dependencies and more — into easily-deployable images that contain everything an application needs to run with increased efficiency across multiple nodes.
Cray System Management Support
To get the most workload production out of your supercomputing system, Cray provides a robust system management stack with a broad variety of utilities and support.
The Cray System Management Workstation leverages industry-standard infrastructure and interfaces to improve ease of use and shorten learning curves, while also providing management innovations to optimize production HPC environment reliability.
To maximize uptime, Cray supports more efficient methods of performing staged system upgrades and rolling out patches while the system remains operational. XC series supercomputers continue to get real work done while intelligently scheduling and managing updates as resources become available and around running jobs.
Cray’s management tools also provide the ability to snapshot the system state and provide protective rollback support, reducing lost operation cycles and downtime managing system updates. The Image Creation and Provisioning Service, Configuration Management Framework and Power Management utilities enable datacenter staff to administer their supercomputers with efficiency and confidence.
Effective System Monitoring Reduces Overhead
The Cray System Snapshot Analyzer (SSA) proactively addresses system health with remote, automated system status checks, or “snapshots.” When failures and issues occur, the snapshots can expedite solution turnaround time. Using the SSA for system query and data collection can dramatically improve issue detection and minimize the typical user overhead involved with communicating symptoms to our support team.