Deadline 30/06/2024
INAF’s USC VIII-Computing issues a new call (the fourth one) for the use of HPC/HTC computing resources and for the availability of data storage spaces.
In particular,
(i) the use of the INAF computing system called PLEIADI will be offered, the technical characteristics of which are described on the site https://pleiadi.readthedocs.io/en/latest/clusters/index.html ,
(ii) the use of the Leonardo BOOSTER computing system will also be available, the technical characteristics of which are described on the site https://wiki.u-gov.it/confluence/display/SCAIUS/UG3.2%3A+LEONARDO+UserGuide,
(iii) the long-term preservation system of scientific products at IA2 will also be available, with characteristics described on the site https://www.ia2.inaf.it/index.php/ia2-services/data-sharing-preservation
Information about these systems for computing and data archiving/preservation can also be found on our website https://usc8.inaf.it
A ticketing system has also been setup to deal with technical problems related with the use of these systems: you can send your messages at pleiadi-help@ced.inaf.it
Who can apply
Both individual researchers and teams belonging to research projects, European projects, PRIN, INAF mainstream projects, scientific missions, etc., which require long-term computing and data storage resources, can apply to the call.
Available Resources
Computing (Pleiadi)
The call in question (which corresponds to number 4 for the PLEIADI system) makes a total of 15 Million CORE hours available, which will be usable starting from 01/08/2024. Proposals for the use of computing e-infra that require a minimum of 100,000 CORE hours up to a maximum of 500,000 CORE hours will be evaluated. The assigned computing resources will be available for up to a maximum of 6 months from the aforementioned date of first use.
Computing (Cineca)
As part of the agreement with Cineca, INAF offers 1.5 Million Standard Hours which will be usable from 01/08/2024. Proposals for computing resources requiring a minimum of 12,500 Standard Hours up to a maximum of 125,000 Standard Hours will be evaluated. The assigned computing resources will be available for up to a maximum of 6 months from the aforementioned date of first use.
In the context of this Call, the resources will be used on the Leonardo BOOSTER calculation system (https://wiki.u-gov.it/confluence/display/SCAIUS/UG3.2%3A+LEONARDO+UserGuide): please remember that 1 Standard Hour is equivalent to 1 CORE hour on Leonardo BOOSTER .
Projects can also require up to 4TB of work storage. Beyond these limits, additional disk resources may be requested (subject to verification of actual availability by Cineca). In this case the additional resources will be calculated as Equivalent Standard Hours with the following scheme.
1 TB/year is equivalent to:
● 500 Standard Hours on DRES-ARCHIVE;
● 800 Standard Hours on DRES-FS;
● 800 Standard Hours on $WORK per HPC
More information on CINECA storage systems can be found at https://wiki.u-gov.it/confluence/display/SCAIUS/UG2.5%3A+Data+storage+and+FileSystems
Long Term Storage (IA2)
As in the previous call, INAF provides long-term data storage spaces, even completely free from the need for computing resources. These long-term storage spaces will be made available starting from 01/08/2024. The typical dimensions of the spaces required in each individual proposal are up to 20TB. Their use may continue until the date indicated and justified by the applicant. Since data preservation will take place on Tape Library devices, it is advisable to carefully read the description of the data sharing and preservation services (https://www.ia2.inaf.it/index.php/ia2-services/data-sharing-preservation ) offered by IA2 as well as the description of the Long-Term Preservation service (https://www.ia2.inaf.it/index.php/ia2-services/data-sharing-preservation/long-term-preservation) for the preliminary measures necessary for preparing the data for the preservation and subsequent recovery of the data.
Proposal submission
Proposals must be submitted by completing the form available at the link: https://forms.gle/MwmL8CcKucFzwtxA8 starting from 01/06/2024 and by 11.59pm on 30/06/2024 .
In the application, the proposing research groups will be asked to specify in detail the scientific background, the technical characteristics of the code, detailing libraries, computing environment, compilers, paradigm and degree of parallelism. Furthermore, you will be asked to describe any constraints on how resources are used (for example dedicated nodes, minimum number of nodes required per run, execution time of a single run, total memory for a single job,….etc) and to indicate the period during which the calculation time will be used.
Proposers will also be asked to specify the storage space necessary for the execution of the code. This storage, not subject to backup, will be available for up to 6 months from the end of the project, and is to be understood as functional for data production (i.e. it does not correspond to preservation storage). If long-term saving of the data produced is necessary, it will be necessary to submit a further and separate request, again via the proposal submission form, for the saving space on Tape Library IA2.
Allocated time that is not used in the requested period cannot be recovered later.
Similarly, the application for long-term data preservation space must contain all the data necessary for the evaluation of the request: size of preservation storage space requested; type, format and size of each scientific product; description of the structure of the collection; description of the expected frequency of access as well as the data access policy; any future plans for publishing the data; all information deemed relevant for correct data preservation following the FAIR principles.
During the proposal evaluation phase, the Panel may request clarifications and further information from the proposers. Communication of any allocation of the requested resources will normally take place within 4-5 weeks of the closing of the call.
Additional info and Ex-post activities
Starting from 01/07/2024 (the next day after the closing of the call), it will also be possible to request computing resources on a “first come, first serve basis”, up to a maximum of 100,000 CORE hours per project, and until the resources available for this methodology, equal to 5,000,000 CORE hours, are exhausted
As regards Cineca and Leonardo BOOSTER, it will be possible to request hours for “tests” up to a maximum of 10,000 Standard Hours per project.
The computing resources assigned in this mode will be available for up to a maximum of 2 months from the date of first use. These proposals will also be subject to an assessment of adequacy and feasibility by USC VIII and/or the management team of the aforementioned computing resources, with response times normally less than a week.
At the end of the project, the proponent will be asked for a short report on the results obtained and the critical issues encountered using the assigned resources. Failure to send this report precludes participation in the subsequent call.