HPC 2014-15 | Soſtware tools
on a target system. As a result, vendors of computation-based products and services can dramatically reduce costs and time-to- market. TTG Apptimizer is available in CUDA
for MS Windows and Linux used with GPUs from Nvidia and in OpenCL for GPUs from Intel and AMD.
www.ttgLabs.com
and users through a single tool, and can thus concentrate on administrating the platform instead of supporting users. Finally, managers can follow the usage of resources and efficiently bill the business units. Te goal is to enable everyone involved in HPC simulations to spend less time dealing with the complexity of such systems. Currently, existing tools are either soſtware- vendor specific or way complex to tackle the vast heterogeneity of the tools used by engineers. Te next step for the SysFera-DS solution will be to provide a way to integrate complex workflows of applications so as to further automate the whole processing chain.
www.sysfera.com
Today, efficient soſtware optimisation is a major challenge for the developers of resource-demanding applications running on hybrid systems (CPU+GPU). Based on a breakthrough soſtware autotuning technology developed by ttgLabs, the TTG Apptimizer toolkit enables all available processing units of hybrid systems to be used simultaneously and provides a load balancing between them. Tis middleware accomplishes several dynamic optimisation procedures, thus endowing an application with the ability to speed up itself in the runtime by automatic tailoring to the ‘hardware + data structure’ bundle. With TTG Apptimizer, developers can easily optimise resource-demanding applications on hybrid platforms and, with minimal recoding and time spent, achieve the highest application performance available
34
X-ISS, a provider of HPC management solutions based in Houston, Texas, developed DecisionHPC soſtware-as-a- service for the operators of large HPC and Big Data systems who need smart insights into cluster operations. DecisionHPC is a reporting and analytics package that delivers business insights into world-wide cluster usage via a single dashboard, so managers can keep operations running efficiently. Te dashboard presents vital KPIs on every aspect of cluster operations including clusters, network, storage, and scheduling. Both current and historical data can be viewed and analysed to perform corrective actions, or for planning.
Univa, the Data Center Automation Company, is a leading provider of automation and management soſtware for computational and big data infrastructures. Its products and global enterprise support give customers the power to manage all their compute resources, no matter how big or where deployed. Univa Grid Engine soſtware enhances
customers’ ability to maximise the value of their existing compute resources by distributing workload and optimising limited application licences, ultimately reducing the time to market for new products.
DecisionHPC delivers a fast ROI. Te
soſtware increases HPC and Big Data profitability by tracking system operating costs related to users and projects, allowing improved chargebacks and quoting. Cluster productivity is also improved by identifying and improving inefficiencies in system utilisation. Analytical insights help HPC operators
make smarter decisions when considering system expansion thanks to the ability to analyse performance trends and project future capacity with DecisionHPC.
www.x-iss.com
Univa Grid Engine soſtware has
functionality for improved scalability, performance, and resource control for small and large clusters. Linux control groups limit resource usage and provide accounting, as well as control of managed resources. Also, the ability to improve scalability for all job queries and client requests, while maintaining scheduler efficiency, has been accomplished by adding a Qmaster read-only thread pool. Finally, Univa Grid Engine natively supports Microsoſt Windows operating systems, no emulation or compatibility libraries required.
www.univa.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 |
Page 39 |
Page 40