If you plan to use or extend the usage of HPC resources for your research activities or if you plan to purchase computing power for your research work, we strongly recommend getting in touch with us at the very early stage of your project. Our role is to provide advices, information and eventually help in preparing the HPC part involved in your projects or grant requests.
If you feel that your computations are not progressing fast enough or feel that you won't obtain the expected results in time, we could help you in improving your workflow or optimising your calculations.
A simple meeting can help in solving many issues. Feel free therefore to contact the HPC team for consulting and working together in preparing and running your research activities.

Information

Before writing grant requests involving HPC, make sure you have all the required information at hand before submitting it. Some funding agencies will refuse to financially support custom-made HPC infrastructures and will favour the use of a centralised/institutional one. Some reviewers might reject a grant request if the technological/technical part(s) of the computing work is not correct, too vague or incomplete.
Contact us for reviewing your HPC related grant requests, early enough, and we'll make sure it will be as complete and correct as possible.

Writing

Formulating properly technical requirements or computing workflows in grant/project documents is essential in HPC-based research too. Grant or projects can be rejected if the processes described in the documents are not valid, sometimes for just one detail, but a critical one.
Please get in touch with us to get help in writing the technical/technological and workflow design parts!

Technological studies

There is a thigh relationship between the type of computational work and the hardware behind it to run the code. Planning the research work must be coupled with planning the hardware purchase or use. Does your code is purely floating-point intensive? Can it scale on mult-cores and/or multi-nodes? Memory capacity or speed can be a bottleneck? Are read/write operations a significant part of the workload? There are many questions to be answered when planning computational research work. We can help in answering those questions and even make market sutdies to find a sweet spot between budgets and resources requirement. Once more, get in touch with us, well in advance, to discuss the technological questions.

Budgets

Last but not least: budget sizing. In several occasions we witnessed allocated grants to financially support the researcher(s), lab equipment but (almost) nothing for the computational resources. Then the data come in and the resources offered for free on our HPC platform are not enough to handle the required storage or computational power. Yet, we offer extra paid resources for such extra needs. When budgeting your research project, get in touch with us to make sure you have allocated a sufficient part of the budget for the computational part of the project.