PERSIST - How to Compute in Persistent Memory Systems

PERSIST - How to Compute in Persistent Memory Systems
Date : 25.11.2022

Conventional main memory technologies (DRAM) loose the data they store in case of a reset of the computing system they reside, which might occur, e.g., due to a power outage, some other failure, or a typical shut-down. This leads to the necessity of periodically saving the work done on the system, so that it is not lost in case of a system reset. Next generation computer systems will rely on emerging memory technologies, such as Non-Volatile Memory (NVM), to provide persistence (and address the high computational demands of modern applications). Persistence ensures that upon system’s recovery, the NVM memory will be in the state it was at the time of the crash (as is the case with disks and other types of secondary storage).

PERSIST (How to Compute in Persistent Memory Systems) is a research project that has the ambitious goal to make a significant step towards addressing the question “How does computation change in the NVMM era and what will be the impact of persistent memory on the way we compute?”. The project is funded by the Hellenic Foundation for Research and Innovation (HFRI) under the 2nd Call for HFRI Research Projects to Support Faculty Members & Researchers. It started on May 2022 and it will last for three years.

PERSIST aspires to have significant impact in the following directions:

  • The proper understanding of the functioning of current and future NVM approaches and the capability to formally assess them.
  • The harnessing of the performance challenges of NVM-based recoverable computing (i.e., computing whose state can be restored after recovery from a failure or a reset), and the better conceptualization of its performance characteristics and boundaries.

The proposed research project has the potential of great economic and social impact in Europe as NVM is expected to bring significant improvements on a wide spectrum of systems and applications, including the future Internet, data centers, cloud computing, big data, e-commerce, cyber-physical systems, and many others. It will also strongly impact exa-scale computing and the future of high-performance computing (HPC), as well as storage systems and architectures, and will be game changer for transaction-oriented systems.

CONTACT INFORMATION
Prof. Panagiota Fatourou
Foundation for Research and Technology - Hellas (FORTH)
Institute of Computer Science (ICS)
Tel.: 2810 391727, Email: faturu@ics.forth.gr
URL: www.csd.uoc.gr/~faturu

MORE INFORMATION

COLLABORATING INSTITUTES