ProteusTM

ProteusTM

Automatic transactional memory adaptation

There are many designs and implementations of Transactional Memory, and that is not a coincidence: different ones perform best on different workloads. One cannot expect the developer to be aware of such concerns, as it is even contradictory with the simplicity advocated by the use of the transactional abstraction. With ProteusTM, the developer writes transactions and the underlying implementation uses techniques to automatically change Transactional Memory algorithms, parallelism degree, and others.

ConcurrencyDevelopment
Key facts
Maturity
Support
C4DT
Inactive
Lab
Unknown
  • Technical
  • Research papers

Scalable Computing Systems Laboratory

Scalable Computing Systems Laboratory
Anne-Marie Kermarrec

Prof. Anne-Marie Kermarrec

Computing systems that make human sense of big data are now ubiquitous. Equipped with powerful AI algorithms, they are now present in all aspects of our life: they drive cars, do surgery, control the lighting in your home, recommend movies and books, and are even about to replace banks. Beyond the AI algorithms and the availability of large volume of data, this is now possible because we can scale systems to thousand, even millions of distributed entities. Yet, designing efficient distributed systems come with many challenges that we address in our team. We are interested in all aspects of such large-scale distributed systems be they datacenters, edge computing, fully decentralized systems, self organizing systems, and we are working on scalable design, failure resilience, performance and privacy-preservation.

This page was last edited on 2022-07-07.