Please consult the following sections to find information about the institute’s projects and resources.
If you are unable to get a suitable answer, or if you have a specific request, contact the institute coordinator.
The institute for computing and data sciences (ISCD) was established in July 2010 to enable researchers to capitalize on the expertise in high-performance computing and data analysis. ISCD is supported by an effort initiated by UPMC with the goal to advance the scientific discovery in the many disciplines brought together around scientific computing and data analysis projects.
The mission of ISCD is multidisciplinary, and involves researchers beyond single areas of research. ISCD provides support for several research groups (project-teams); co-location of faculty, engineers and graduate students;semnar series; summer schools; development of open source tools; and outreach efforts.
ISCD programmes support and increase the community of researchers engaged in HPC and data analysis research so that they can advance software tools that can be translated across research disciplines, and devote more time to their research; contribute to new training opportunities; and share best practices across disciplines. The initial programmes providing funding to scientists are listed here.
There are several ways of getting involved in ISCD programmes. If you are a member of the UPMC-SU community (faculty, staff or graduate student), please consider submitting information about your field of expertise and research topics. This will help us to identify experts and to update a portfolio of computing and data sciences activities at UPMC.
If you have general question/suggestions, write to firstname.lastname@example.org.
If you are an industry representative and would like to engage in collaboration with ISCD researchers, send a brief description to email@example.com.
The initial programmes supported by the institute for computing and data sciences are described here.
Currently, faculty and staff may apply to become associate members of the project-teams. The team leader(s) will validate the application based on scientific expertise and complementarity with the other scientists of the project-team.
Yes. Send an email to firstname.lastname@example.org describing your question and the type of expertise you need.
The research model of the institute for computing and data sciences is founded on a basic component, the project-team, that structures the institute’s activities. Each team is composed of a dozen researchers and staff and is led by one or a pair of scientific personalities who define the scientific objectives. Find here the active project-teams of ISCD.
An overview of the various research topics can be found on our research pages.
The most straightforward way of getting involved in a research project-team activities is contacting its scientific leader(s). ISCD offers research opportunities for graduate and undergraduate students alike. Check the web pages of our project-teams for more details.
We have job listings and research/support openings on our news page.
Yes. Our summer internship programme is designed to provide undergraduate students (levels L3-M2) an opportunity to work and study with some leading scientists in an environment devoted to scientific research: the marine station in Roscoff. Check the web page for more info.
Yes. Every student currently enrolled in a university program (level L3-M2) is eligible to apply to this program.
No. However, most summer school topics are in scientific computing and data analysis with a focus on applications. Check for the prerequisites in the booklet.
Yes, the application deadline is May 15. Note: partial applications that are not completed by the May 15 deadline will not receive any further consideration.
We recommend that you apply as soon as possible after the application site becomes available (early March), as acceptances are made on a rolling basis (and their curriculum). Applicants who submit their materials during the first 3 weeks have about 2 times as likely to be accepted in the programme as those who submit during the 2 weeks just before the deadline.
No. You will need to reapply and request new letters from your references. Note: priority is given to new students applying for the first time to the programme.
We receive over 60 applications. Approximately 30 students are hired each summer. Priority is given to students attending the whole programme (4 weeks).
No. Unfortunately, due to the large volume of applications received each year, only the candidates selected for the programme will be informed. All candidates selected will have been contacted by mid-June.
A computing cluster is a set of high speed computers for intensive computations and data processing. The HPC systems include several hundreds processors (thousands of cores) that can work concurrently on very large problems. More information and a complete description of the current MeSU systems can be found by visiting the resources page.
If you have never used such system, or are not familiar with, keep in mind that these are very exotic machines by virtue of the technologies they are built of. Supercomputers, like sport cars require large amounts of money and specialized expertise to use. They excel only for specific problems. The point of using a HPC system is that the processors work together to solve a problem larger than any one computer can solve. However, the processors (or nodes) have to talk to one another and share information, just like people, in order to meaningfully cooperate. In addition, HPC clusters won’t run without a dedicated operating system; a dedicated version of Linux in our case.
To help beginners, free Level 0 support is provided by the HPC support team.
The HPC MeSU centre provides high-performance computing (HPC) resources at no direct cost to UPMC-SU faculty and staff, for research purposes. These resources include two HPE-SGI Linux-based systems, software modules (compilers, applications), and commodity services (ethernet access, high-speed scratch storage, security).
If you are a PhD, postdoc, faculty or staff member of UPMC-Sorbonne University engaged in university-related research you are entitled to an HPC account, just by filling in a HPC Account Application form. Other academic/non-academic researchers and engineers are welcome to an account also.
If you have any question about your eligibility for an account, please contact email@example.com with questions.
Check this page and follow the instructions to get started.
Yes. To request access to HPC resources, consult the account subscription services.
MeSU systems run SuSE Linux Enterprise Server (SLES 11 SP2), a Linux-based operating system.
User data is kept isolated from others. With our Linux systems, another user cannot see nor access your program and data. IT experts insure data integrity through various security measures.
Since 2017, ISCD is rolling out the MeSU Computing Test (MDT), a software skills aptitude test for researchers wanting to use MeSU’s HPC resources. The MeSU computing test has been developed by the MeSU Support team.
This test consists of practical questions that cover both MeSU’s resources (eg the type of processors, the operating system) as well as software development and running within MeSU’s infrastructure (eg the task scheduling system, the monitoring of resources). The test is taken online by potential users, and takes about 1hr max under the supervision of a trainer of the support team. Feedback is provided about both correct and incorrect answers, with possible links to further information.
The test is optional for experienced users (with experience on other HPC systems), but is mandatory for beginners and newcomers. Users who pass the test are given time on MeSU as a reward, to encourage them to apply some of their new learning immediately.
|User categories||CPU hourly rate||Storage
|UPMC-SU||free of charge||included||included|
|other academics||0.03 €/core-hour||included||included|
|SME/ETI, companies||0.05 €/core-hour||included||included|
The rates shown here correspond to an hourly rate per core. Each hour partially consumed is charged as a full hour. A report of the balance credits is produced every three months or at the end of the project.
There is no charge for time on the login nodes.