extreme factory

extreme factory_620

HPC-as-a-Service from Bull: innovation at your fingertips

Computer simulation has become an essential tool for innovation. But a lack of resources or expertise means many companies and organizations do not have access to an HPC infrastructure. extreme factory is a set of Cloud services combined with HPC expertise that helps these organizations implement their computer simulation applications. Born out of Bull’s unique experience in HPC, extreme factory gives businesses of all sizes, in all sectors, access to made-to-measure supercomputing resources.

Flexibility, performance, security, service

To meet the wide range of needs, extreme factory is available in four configurations: on-demand pay-per-use, on-demande reserved, private on premises and private hosted. It’s also possible to combine these configurations into hybrid solutions. Whatever arrangement you choose, it is implemented and operated by Atos experts. extreme factory offers superior levels of performance, reliability, availability and security, and benefits from all the power and wealth of environments available on Bull’s world-class supercomputers.

Using exclusive tools like the extreme factory computing studio portal and extreme factory remote visualizer 3D visualizer, users can access a complete work environment and personalized service that allows them to conveniently manage every stage of their project: from data loading to visualizing the results. The flexibility and elasticity of the Cloud means everyone can have the resources they need; simply, cost-effectively and securely. System administrators also benefit from greater usability, because the portal lets them manage resources, tasks and users centrally.

Keeping the cost of innovation under control

As well as the advantages of sharing infrastructures, optimizing how they are used and rationalizing of licenses, extreme factory boosts productivity and gives you total visibility of expenditure by project, department or resource. It’s easy to establish usage indicators and detailed reporting, for accurate cost allocation. With extreme factory, you can keep the cost of innovation under control.

Developed in partnership with numerous ISVs, extreme factory is compatible with most commercial and Open Source applications. So extreme factory meets the demands of all sectors, from aerospace, automotive, engineering and chemical industries to animation, video, games and financial services. An integrated service offering a unique combination of technology and expertise, extreme factory puts innovation at everyone’s fingertips.

 

Four HPC-as-a-Service packages for a made-to-measure solution
extreme factory on demand (XFoD) extreme factory private (XF private)
XFoD pay-per-use >> XF private on premises >>
XFoD reserved >> XF private hosted >>

 

extreme factory

extreme factory on demand pay-per-use

All the simplicity of pay-per-use

extreme factory on demand gives you access to pooled computing resources, hosted and operated by Bull, and billed in real time as you use them. Users connect to their customized work environments on the Internet via a secure portal, where they can launch their simulations on demand. The processing cores are shared between different customers, so task scheduling is on a ‘first come, first served’ basis. This simple, cost-effective and flexible option is ideal for impromptu needs or where there is no great urgency, such as POCs… As part of a hybrid Cloud, it can also provide additional computing power if necessary.

extreme factory on demand reserved

Guaranteed resources at optimum cost

Project imperatives often mean it’s hard to accommodate the uncertainties that come from using shared resources. But given that periods of intense demand (bursts) can usually be anticipated, Bull offers extreme factory on demand in reserved mode. This service allows users to pre-book a processing time-slot. No more risks associated with having to wait: resources are guaranteed and freely available during this time. The reserved time-slot is paid for, regardless of the use you make of it. This formula means you get all the benefits of pooling and the flexibility of the public Cloud, while enjoying guaranteed resources. Use on its own or in addition to a private Cloud in a hybrid configuration, it is particularly suitable for one-off, CPU-hungry projects.

extreme factoy private on premises

Your infrastructure, Bull’s expertise

Do you want to keep your computing and visualization resources within your business, while letting your users benefit from the flexibility and power of the Cloud? With extreme factory private on premises, the Cloud infrastructure stay within your walls and under your operational responsibility. Bull experts are responsible upstream for its definition and implementation: deploying new infrastructures and optimizing existing ones, consolidating and upgrading resources, harmonizing processes, structuring internal offering… Located on your premises, the extreme factory computing studio portal becomes the sole gateway to your private HPC Cloud, facilitating how the computing or visualization environment is used by the business and how it is managed by your system administrators. Your HPC resources are totally integrated, streamlined and secure, to take full advantage from the elasticity of the Cloud.

extreme factory private hosted

Innovate with total peace of mind

Bull offers extreme factory private hosted, where your HPC Cloud infrastructure is designed, implemented, hosted and operated by Bull experts in a secure Data Center. The connection is via secure Internet gateway (at a custom bandwidth) or VPN supported by your operator or Bull. This approach allows for full customization to provide you with the computing environment that best suits your patterns of usage and your applications, while retaining the flexibility of the public Cloud. You get defined service levels and maximum performance without the need for your teams to get involved, so they can focus on other tasks. This model is particularly well suited to companies and organizations that need long-term access to computing resources: extreme factory private hosted lets you concentrate all your energy on your innovations.

Extreme factory frees up your capacity for innovation

A single point of access >>

The extreme factory computing studio secure portal gives users and system administrators alike direct access to all resources and applications. The user interface is born out of a continual dialogue between Bull and its customers, so extreme factory computing studio truly reflects users’ expectations when it comes to HPC applications.

Once a user is logged on, he or she can access a complete work environment, customized for their job role and applications. From there, they can load and manage data, set parameters for the simulation, run the calculation, track its progress, and then proceed to post-processing and visualization of results. Administrators can manage users and roles, resources, task scheduling, application licensing…

As a single gateway to the HPC environment, extreme factory computing studio simplifies the handling and management of an HPC cluster, and lets you focus on the actual calculation and results.

Security: essential at every level >>

Whether it’s for  fundamental research or new product development, the data from computer simulations is among the most critical for companies and laboratories alike, as it goes right to the heart of their investments, strategy and know-how. For Bull, security is a fundamental component of HPC: so our extreme factory solutions natively integrate very high levels of security to protect data and exchanges. Interconnections are covered by a secure protocol, IP addresses are filtered (only the client’s own domain name is allowed), and users can only access a virtual machine, not the physical machine, to ensure total data partitioning. For security reasons, all application installations are carried out by Bull teams. Other options (VPN, encryption…) are possible depending on the context and needs.

A convenient and secure visualization tool >>

Bull’s exclusive 3D visualization module, extreme factory remote visualizer (XRV) is compatible with most available applications and lets you use post-processing modules directly on the server: only the compressed image is transferred to the user workstation. Exchanging less data means greater security, more fluidity, and less use of infrastructure resources (networks, user workstations…), as well as being more convenient for users and offering more opportunities for increased collaboration.

Application expertise and service: getting more from support >>

In areas as sophisticated as industrial design and scientific research, the slightest difference in the quality and accuracy of results can have a significant impact. To gain an extra edge in performance and obtain perfectly understandable and usable results, each link in the chain has to be optimized: and only a person who knows the area of application, the software used, the vocabulary or common way to present the results will be able to translate business needs into technical settings.

In Bull’s dedicated HPC teams, technological expertise is coupled with genuine scientific know-how. Thanks to this dual culture, the services that support extreme factory solutions provide a significant added value at every stage of the project: the design of the solution architecture; platform design and optimization; the ergonomy and securty of the work environment; integration of third-party tools, post-processing of the data produced… In addition – when it comes to  hosted solutions – industry-scale application services (maintenance, monitoring…) ensure high levels of service and usability via a single point of contact.

 

REFERENCES

Exa - private CFD cloud           STX France - HPC on demand

Exa – private CFD cloud      STX France – HPC on demand

 

UberCloud experiments

Discover four case studies, out of a dozen team experiments made with extreme factory resources, from computational fluid dynamics, multi-physics, combustion, and molecular dynamics. They document team results, their findings, challenges, lessons learned, and recommendations.

Fill in the form below to download the white paper