Jump to content

Nir Shavit

From Wikipedia, the free encyclopedia
Nir Shavit
Alma materTechnion,Hebrew University of Jerusalem
Known forSoftware transactional memory,wait-freealgorithms
AwardsGödel prize,Dijkstra prize
Scientific career
FieldsComputer science:concurrentandparallel computing
Thesis(1990)
Websitewww.cs.tau.ac.il/~shanir/

Nir Shavit(Hebrew:ניר שביט) is an Israeli computer scientist. He is a professor in theComputer ScienceDepartment atTel Aviv Universityand a professor of electrical engineering and computer science at theMassachusetts Institute of Technology.

Nir Shavit received B.Sc. and M.Sc. degrees incomputer sciencefrom theTechnion - Israel Institute of Technologyin 1984 and 1986, and a Ph.D. in computer science from theHebrew University of Jerusalemin 1990. Shavit is a co-author of the bookThe Art of Multiprocessor Programming,is a winner of the 2004Gödel Prizein theoretical computer science for his work on applying tools fromalgebraic topologyto model shared memory computability, and a winner of the 2012Dijkstra Prizefor the introduction and first implementation ofsoftware transactional memory.He is a past program chair of the ACMSymposium on Principles of Distributed Computing(PODC) and the ACMSymposium on Parallelism in Algorithms and Architectures(SPAA).

He heads up the Computational Connectomics Group at MIT'sComputer Science and Artificial Intelligence Laboratory,focusing on techniques for designing, implementing, and reasoning about multiprocessors, and in particular the design ofconcurrent data structuresformulti-coremachines.

Recognition[edit]

Currently he has co-founded a company named Neural Magic along with Alexzander Mateev. The company claims to use highly sparse neural networks to make deep learning computationally so efficient that GPUs won't be needed. For certain use cases they claim a speed up of 175x.[2]

References[edit]

  1. ^ACM Names Fellows for Computing Advances that Are Transforming Science and SocietyArchived2014-07-22 at theWayback Machine,Association for Computing Machinery,accessed 2013-12-10.
  2. ^"The Future of Deep Learning is Sparse. - Neural Magic".12 July 2019.

External links[edit]