Intensive Computing

The successful merge of technologies

Get demo

Based on the principle of distributed systems, ActiveEon offers a solution that virtually connects all types of computers, unifies them to create working clusters, and takes advantage of parallel processing to improve business performance.

ActiveEon: The Convergence of Technologies & The Delivery of Unique Features

Cluster Computing

Similar computers are connected through a fast LAN to operate as a single system.

Cluster Computing icon

Grid Computing

Different types of computers are linked together from different locations and can operate independently.

Grid Computing icon

Cloud Cluster

Provide on-demand IT resources and services, often not shareable. A cluster can be created through a third-party tool or Kubernetes.

Cloud icon

High-Performance Computing

Use parallel data processing to improve computing performance and perform complex calculations.

High-Performance Computer icon

ActiveEon creates a new paradigm by taking all the strengths of these technologies and adding more features.
It virtually connects all types of computers, groups them to create clusters, and utilizes parallel data processing to improve computing performance. Additionally, it enables service on-demand with shareable resources.

Hybrid Orchestrator

With ActiveEon, you can deploy and centrally manage all workloads as well as distribute tasks execution in parallel to all relevant computing resources.

These computing resources can be High-Performance Computing, Grid Computing, Cluster computing and all the declination of these technologies delivered by cloud providers as a service (AWS, Azure, GCS...).

Users can unlock silos and benefit from easy connections to all computing resources.

From a single interface, Administrators can manage the end-to-end process across the entire organization infrastructure (on-Premises, in the cloud, at the edge).

ActiveEon - Intensive Computing diagram

Learn more

Any questions? Feel free to get in touch with us, we would be happy to have a chat with you and help you with your project!

Contact Us

Meta Scheduler

Easy connection illustration

Easy connection

As the technology evolves and becomes available anywhere in the cloud, HPC technology needs to become accessible to a broader audience. With ActiveEon, you can connect seamlessly to existing HPC native schedulers (preserving past investment from HPC customers) and give new users access to “Intensive computing” at a press of a button from a modern web interface.

Adaptability illustration

Adaptability

Enterprises and end-users now have the flexibility to run their workload in the most suitable environment based on technical or business consideration. Some users will let the meta scheduler pick the first available machine, while others can select a specific set of machines ideally suited to their type of workload.

Flexibility illustration

Flexibility

ActiveEon workflows have the unique capabilities to be completely independent across all computing platforms – saving users time and making automation much easier across the enterprise. F or example, by using a single ActiveEon workflow, end-users can process part of their workload in a locally distributed environment, part in several cloud VMs, and part using the HPC systems. The choice is yours, and if you do not make any specific choices, the ActiveEon algorithm will distribute the processes in parallel looking for available cores.

This powerful feature is ideal for AI and ML pipelines but also for many Big Data pipelines where the technology stack becomes large and complex and where users are requesting to orchestrate and automate as much as possible using the resources that are the most adequate.

ActiveEon: Optimize your Resources

Computing Resources icon

Connect Computing Resources

ActiveEon's solution virtually unifies different type of computing resources together.

Work together icon

Make them work together

The unified compute resources are managed and used as a virtual supercomputer.

Local to remote location icon

Local to remote location

Single computing resources can be located anywhere: on-premises, in the cloud, or in an HPC.

distributed computing icon

Distributed Computing

Our advanced distributed computing technique divides the workloads into tasks and assigns them to each available CPUs and GPUs in parallel.

compute optimization icon

Compute optimization

The solution optimizes the use of resources by constantly monitoring all CPUs and GPUs and distributing tasks to the first available core.

Easy Provisioning icon

Easy Provisioning

Scale your workload with Easy Provisioning, where users can spin up new resources at a press of a button.

Easy Scaling icon

Easy Scaling

With dynamic bursting, the solution will automatically spin new compute resources so that time-constraint workloads can be executed on time.

Get a demo

Our team would be glad to help you understand how our solution allows you to benefit from the merging of intensive computing technologies!

Get demo

High-Performance Cloud Clusters

When end-users launch workloads on multiple applications, ActiveEon will automatically provision cloud instances and combine cloud instances to make them to work together.

Cluster together computing resources and use them as a virtual supercomputer

To resolve complex computational problems and manage intricate tasks priorities, end users will use ActiveEon to aggregate the several cloud instances to work as a single entity and use them as a virtual supercomputer.

The solution will distribute tasks automatically and launch them in parallel. Additionally, the solution at all time monitors every core within computers and VMs and distribute tasks as soon as a core is available.

The solution also manages complex job priorities that are important for some workloads such that some particular tasks must run prior to others; and some tasks might require particular server configurations in terms of memory, CPUs, GPUs or other particular hardware on which to run.

Workload balancing Clusters

Improved performances

The solution distributes the workload amongst multiple instances to improve the performances. The load balancing facility has been designed to avoid overstressing any single instance by splitting up the jobs or traffic flow.

Task distribution

Instead of simple round-robin distribution of workloads, ActiveEon actively monitors all cores and when a resource is available distributes tasks according to a set of priorities.

Download our use case

Learn more about how our solution has been a game changer for Legal & General by significantly speeding up their risk analysis!

Download Use Case

Easy provisioning and Dynamic burst

Additionally, to create a powerful shared “intensive compute” environment, ActiveEon enables users to spin up on demand high-performance clusters quickly and efficiently using containers or VM technology.

The request for new compute resources could be initiated from a simple web interface or from your enterprise ITSM solutions.

For the users or departments who have the right to do so, ActiveEon also provides a technical solution to those short-term peaks. It enables the dynamic spin up of additional instances when workloads have critical time constraint. At the end of the process, the solution will automatically spin down the instance to protect your cloud budget.

ActiveEon: More control, better features

Target icon

Select compute targets

Users can select the best computing machine to process their job.

Experts can easily select one or a group of fast machines to process workloads that require speed.

The resource targets can be physical or virtual machines.

Data Scientists can select from their laptops, remote HPC clusters or cloud HPC-As-A-Service for powerful model training.

ActiveEon democratizes the use of HPC by giving users an easy-to-use modern web user interface.

Experts can easily launch workloads and ActiveEon will communicate seamlessly with the native HPC scheduler.

testing icon

Real-time testing using different parameters

Users can test their workloads on different compute targets and then select the one that guarantees a quicker and better result.

Users can select a simulation program and test its efficiency on different computing resources to validate the best production environment.

Interactive icon

Interactive Job Control

Allow a user to interact with the workload in real time within any environment.

Once a workload is launched, Experts can modify real-time the parameters to orient in a different direction the behavior of the simulation.

CPU icon

CPU-Intensive

ActiveEon optimizes CPU-intensive tasks across the environment, using distributed computing, task queuing, and intelligent job priorities.

When launching CPU-Intensive workloads, experts know that we will manage the distribution of tasks across the environment, prevent the shutdown of resources and prioritize important tasks.

RAM icon

RAM-intensive

Like for CPU-Intensive, users with RAM-intensive needs can select a single of a group of machines adapted for their type of workload.

DataOps users can easily create virtual groups of machines that have at least 16 GB of RAM and select that cluster to crunch the data.

central operation to remote data icon

From central operation to remote data crunching

Thanks to ActiveEon universal agents, the central scheduler communicates with remote universal agents and processes workloads on the selected remote machines.

An expert can use their laptop to set up a simulation and select the group of machines close to where the data is located. The scheduler will communicate with universal agents and the data is processed in the selected remote machines.
Essential for avoiding the transfer of a vast amount of data through the network or the need to keep the data within a certain geography.

self-Healing icon

Self-Healing Resource Failure

ActiveEon monitors at all times every resource and will detect any machine failure. In a case of a failure, ActiveEon automatically relaunches the loss tasks and completing the workload.

Administrators can rely on the ability of the solution to detect and remediate issues without human intervention. The solution, as well as finishing the workload, will alert IT professionals of the loss of resources, giving them a quick way to identify the root cause.

Converging "Intensive Computing" technologies

With the ActiveEon’s solution, you are able to lead the way in creating a hybrid architecture, where existing silos are disappearing and where data and applications can be easily shared using any type of compute resources.

With your unique scheduling platform, you can unify and converge all compute resources, automate the load on CPUs and GPUs to optimize the time sensitive workloads and control the end-to-end process.

The various applications used by “intensive computing” will be managed on ActiveEon unified platform, interconnecting data assets to the various other applications used across the enterprises.

Data silos will be broken down, ensuring that no data will require to be migrated, keeping unchanged the data foundation and the applications, thus optimizing TCO and boosting ROI.

Go Further

Photo of business team sharing info
Get in touch

Feel free to get in touch with us to learn more.

CONTACT US
Photo of a business expert sharing info with client on computer
Talk to an expert

Get in touch with one of our experts.

Talk to an expert
Photo of working man with computer
Get the use case

Read about how migrating to ActiveEon is a game changer.

Download