Modern scientific software requires ever-increasing computing capacity, necessitating increasingly specialized systems to run on purpose-built supercomputers. Given that each supercomputer is generally a one-off project, the need for computing frameworks portable across different CPU and GPU architectures without performance losses is increasingly compelling. Pennycook’s portability score allows for quantifying how performance varies across HW platforms. We use it to study the properties of different HPC programs, for example, the solver module of the AVU-GSR pipeline for the ESA Gaia mission. This code finds the astrometric parameters of 108 stars in the Milky Way using the LSQR iterative algorithm. Besides its astrophysical relevance, LSQR is widely used to solve linear systems of equations across a wide range of high-performance computing applications. We optimized the previous CUDA implementation and ported the code to further six GPU-acceleration frameworks: C++ PSTL, SYCL, OpenMP, HIP, KOKKOS, and OpenACC. Thanks to the great availability of heterogeneous resources on HPC4AI, we evaluated each framework’s performance portability in terms of application and architectural efficiency across different GPUs. Specifically, we use NVIDIA T4, V100, A100, and the new GH200. For each GPU, we compute the roofline model to get the architectural efficiency.

Contacts:
Contact: Giulio Malenza

High-throughput sequencing Chromosome Conformation Capture (Hi-C) allows the study of chromatin interactions and 3D chromosome folding on a larger scale. A graph-based multi-level representation of Hi-C data is essential for proper visualisation of the spatial pattern they represent, in particular for comparing different experiments or for re-mapping omics-data in a space-aware context. This project aims at developing NeoHiC, a user-friendly web application for the progressive graph visualisation of Hi-C data based on the use of the Neo4j graph database.
Contacts:
Daniele D’Agostino – dagostino@ieiit.cnr.it
Source code:
https://github.com/dddagostino/neohic

This project targets two broad problems:
– the former has to do with the contextual nature of lexical meaning, which is at the base of semantic composition;
– the latter involves extracting and building systems and resources that grasp verbal semantics.
Various sorts of embeddings and neural architectures will be explored, and various applicative domains (amongst which documents categorization, keywords extraction, open domain question answering, text summarization, etc.) will be considered. Further information on broader scientific interests and recent work can be found at http://ls.di.unito.it.


Contacts:
Daniele Paolo Radicioni
e-mail: daniele.radicioni@unito.it

The ALaN platform assists the researchers in different laboratory activities. Its modular architecture allows managing different kinds of raw data (e.g., biological, molecular) and tracking experimental data. Moreover, the platform allows the integration of different resources and supports heterogeneous analyses to improve the knowledge about tumors. User interfaces are designed to be practical in hostile environments, in which researchers should minimize their interactions with the system during data entry procedures (e.g., in sterile conditions).

Contacts:
Andrea Bertotti – andrea.bertotti@unito.it