PAStor

PAStor’s work published at USENIX FAST and IEEE/ACM CCGrid

2022-01-02T00:00:00+00:00

The worked done on the scope of PAStor was published at two renowned systems research conferences.

The paper PAIO: General, Portable I/O Optimizations With Minor Application Modifications was published at the USENIX FAST’22 conference.

The paper Accelerating Deep Learning Training Through Transparent Storage Tiering was published at the CCGrid’22 conference.

Virtual e-poster at UT AUstin Portugal 2021 Annual Conference

2021-10-31T00:00:00+00:00

A Virtual e-poster of PAStor was presented at the UT AUstin Portugal 2021 Annual Conference. Available here.

Two new papers published at the REX-IO Workshop.

2021-10-21T00:00:00+00:00

Two new papers describing the outputs of the PAStor project were published and presented at the Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO), colocated with the Cluster conference.

PAStor project presented at online webinar

2021-09-16T00:00:00+00:00

The research outputs of the PAStor project were presented at the Webinar Series On the Road to HPC: Major Challenges and New Opportunities.

The title of the talk is: Is HPC ready for “Big” Data Storage?, and its video recording can be checked here.

PAStor project talk at Encontro Ciência’21

2021-06-30T00:00:00+00:00

The research challenges and goals for the PAStor project, which will improve the convergence between AI applications and HPC supercomputers, were discussed at Encontro Ciência’21 in Lisbon, Portugal.

Namely, the talk was given at the UT Austin Portugal Program’s workshop with the title: “Better prepared: Strengthening societies through international scientific collaboration”.

PAStor researchers proposes a trusted hardware-based privacy-preserving deduplication system

2021-05-21T00:00:00+00:00

INESC TEC’s team has published a paper entitled “S2Dedup: SGX-enabled Secure Deduplication” at the 14th ACM International Systems and Storage Conference (SYSTOR ’21), as part of PAStor project, which took place online, between June 14 and 16, 2021. This paper was also selected as best paper runner-up to this conference.

This publication proposes a trusted hardware-based privacy-preserving deduplication system. S2Dedup was designed to support multiple security schemes, which enables tailored deployments for applications with distinct requirements in terms of security, performance, and space savings.

The authors of this paper are Mariana Miranda, Tânia Esteves, and João Paulo, from INESC TEC and University of Minho, and Bernardo Portela, from NOVA LINCS and University of Porto.

HPBD’21

2021-04-27T00:00:00+00:00

The 1st Workshop on High-Performance and Reliable Big Data (HPBD’21) will take place on September 20th, 2021, in Chicago, USA and is co-located with SRDS 2021.

The HPBD’21 workshop is supported by PAStor and BigHPC projects

For more information please visit https://hpbd-21.github.io.

INESC TEC and AIST collaboration

2020-12-16T00:00:00+00:00

INESC TEC researchers started collaborating with researchers of AIST, in Tsukuba, Japan. This initiative focuses on some of the research topics that the team is currently working on, regarding the PAStor project.

If you are interested, click here to read the researchers’ statement.

Kickoff Meeting

2020-10-09T00:00:00+00:00

The kickoff meeting of the PAStor project took place on October 9 2020 via a virtual platform.

The main objective of the project is providing a novel Software-Defined Storage (SDS) solution for HPC that can efficiently handle I/O flows from multiple AI workloads by automatically adjusting storage configurations and resources to meet applications dynamic requirements.

By gathering the expertise of INESC TEC and Hood College researchers in the AI and distributed storage fields, and by including researchers from TACC and MACC with experience on managing HPC infrastructures, the project will produce new high quality research findings and advance the state-of-the-art for storage solutions currently deployed at HPC centers.