HIGH PERFORMANCE COMPUTING
Tech focus: software tools
A ROUND-UP OF THE LATEST SOFTWARE TOOLS FOR SCIENTISTS AND RESEARCHERS USING HPC
As High-performance computing becomes a more ubiquitous tool
for scientists and researchers in their daily lives, ensuring ease of use of these systems is important to accelerate productivity and scientific discovery. Tools that help scientists manage their workloads and resources more effectively can help reduce the barrier to entry for new organisations or inexperienced users. Helping scientists push their workloads into the cloud, maximising the use of in-house resources or optimising the codes and increasing the efficiency are just some of the examples of benefits that help drive productivity and increase scientific output. Debugging and code optimisation tools can help to reduce issues that may arise from software development. This has benefits in reducing the cost of code development and gives tools to scientific domain experts that help them better implement their scientific research to applications. This frees up time of expert staff and makes it easier for companies to pursue their computing research or product development, without investing as heavily in with coding and application development. Making software tools accessible also helps to
8 Scientific Computing World Autumn 2020
reduce the failure of workloads and helps management teams understand what has gone wrong when things fail. Whether it be hardware monitoring delivered in cluster management software or job scheduling services which report on workload performance, HPC teams can use this information to more accurately support users and deliver a more robust HPC service for scientists and engineers.
Commercial products
Adaptive Computing’s Moab HPC Suite is a workload and resource orchestration platform that automates the scheduling, managing, monitoring and reporting of HPC workloads on a massive scale. Its patented intelligence engine uses multi-dimensional policies and advanced future modelling to optimise workload start and run times on diverse resources. Nodus Cloud OS is intelligent cloud management technology that enables seamless access to compute resources, whether on-premise or in the cloud. Cloud deployment on any of the leading cloud providers becomes easily attainable with Nodus. It simplifies the complexities of running HPC and enterprise workloads and applications in the cloud.
Altair provides high-throughput computing solutions that help you orchestrate, visualise, optimise and analyse in the data centre and the cloud. The Altair PBS Works suite includes workload manager Altair PBS Professional, designed to improve productivity, optimise
utilisation and efficiency, and simplify administration for clusters, clouds and supercomputers, from the biggest HPC workloads to millions of small, high- throughput jobs.
Appentra’s Parallelware Trainer and Parallelware Analyser are novel tools for the development of C/C++/Fortran parallel code for multicore CPUs and GPUs using OpenMP and OpenACC. Designed in collaboration with experts in parallel programming from HPC, they provide a systematic, more predictable approach that leverages parallel programming best practices, and enables writing parallel codes performing to as high a standard as those written by experts.
Arm Allinea Studio is a complete suite of tools for developing Arm-based server and HPC applications and Arm Neoverse processors. It provides commercially supported
”Debugging and code optimisation tools can help to reduce issues that may arise from software development”
products for simple and efficient code generation, and full capability for understanding and optimising workloads for optimum performance on Arm.
Bright computing’s Bright Cluster Manager software offers an integrated solution for building and managing HPC clusters that reduces complexity, accelerates time to value and provides enormous flexibility.
Bright Clusters for Hybrid Cloud automates the process of building and managing a Linux cluster.
Charmworks’ Charm++ is a mature, highly scalable parallel programming framework that enables applications to effectively scale regardless of the execution landscape. Charm++ provides a C++ type of interface that works with Fortran, C and C++ codes. Charm++ helps creators of compute-intensive software deliver better shared and distributed memory parallel application performance.
TrinityX, from Clustervision, is designed to provide all services required in a modern HPC system, allowing full customisation of your cluster installation. With TrinityX, built-in services
like OpenLDAP, SLURM, Zabbix, NTP and many more
@scwmagazine |
www.scientific-computing.com
ZinetroN/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38
