Parallelize your workloads with an open solution
Sep 19, 2018 from Activeeon
The principle is simple: large task can be divided into smaller problems which then solved in parallel at the same time. For instance, in parallel computing applied to chemistry, 10000 computers can test 10000 chemical components in one hour simultaneously. Nowadays, as power consumption and heat generation are an important concern, parallel computing has become the dominant paradigm in computer architecture, very often taking form of multi-core processors.
Parallel computing is used in:
Activeeon solution provides a uniform parallel computing interface with distributed and parallel workflows, and a uniform resource management, independent from the underlying virtualized infrastructure, for better utilization of existing resources from desktop, multi-cores, servers, clusters to grids and clouds.
ProActive Workflows & Scheduling integrates with de-facto standards in scientific and engineering environments such as Matlab or R. Directly from within these familiar environments, it provides users with the capacity to parallelize and distribute executions and manage data transfers on any computing resources: desktop, local clusters, hybrid clouds or multi-cloud.
Parallel computing will save your time and money. More resources working to solve one task will shorten the time of its execution and generate savings. It will be the best solution for larger and more complex problems that are impossible to be solved on a single computer. You will be able to manage multiple compute resources to do many tasks simultaneously. This structure will let you use compute resources on a wide area network, even when local compute resources are scarce or insufficient.
Parallel computing solutions from Activeeon provide a 10-days trial version. Get a free access to our ProActive Cloud platform and all our tutorials helping you to do your first steps on our plateform. While starting to crete your own virtual environment, you will benefit from a free support of our engineers during 10-days period.
Apr 25, 2019 from Nicolas Narbais
Auto ML theories are gaining popularity and multiple solutions are gaining traction in the open source community. However, there is still a large gap between theory and practice. Let's identify some of the challenges....