Multicore Computing Group

We design and investigate the algorithmic aspects of systems with multiple computing elements, in particular tightly coupled multicore machines. As multicore machines become our mainstream computing platform, their unique architectural properties are forcing us to rethink our basic algorithms. The goal of our group's research is to investigate how to use the new properties of the architectures to improve performance, while finding ways to overcome their limitations.

Faculty

PhD Students

MSc Students

Recent Alumni

Recent News

• Yehuda Afek, Amir Levy and Adam Morrison's paper Software-Improved Hardware Lock Elision is accepted to PODC 2014.

• Alex Matveev and Nir Shavit's paper Reduced Hardware NOREC: An Opaque Obstruction-Free and Privatizing HyTM is accepted to TRANSACT 2014.

• Hillel Avni and Bradley Kuszmaul's paper Improving HTM Scaling with Consistency-Oblivious Programming is accepted to TRANSACT 2014.

• Adam Morrison and Yehuda Afek's paper Fence-Free Work Stealing on Bounded TSO Processors is accepted to ASPLOS 2014.

• Yehuda Afek, Anat Bremler-Barr and Liron Schiff's paper Recursive Design of Hardware Priority Queues is accepted to SPAA 2013.

• Alex Matveev and Nir Shavit's paper Reduced Hardware Transactions: a New Approach to Hybrid Transactional Memory is accepted to SPAA 2013.

• Hillel Avni, Nir Shavit and Adi Suissa's paper Leaplist: Lessons Learned in Designing TM-Supported Range Queries is accepted to PODC 2013.

• Alex Matveev and Nir Shavit's paper Reduced Hardware Transactions: a New Approach to Hybrid Transactional Memory is accepted to TRANSACT 2013.

• Hillel Avni, Nir Shavit and Adi Suissa's paper Leaplist: Lessons Learned in Designing TM-Supported Range Queries is accepted to TRANSACT 2013.

• Join us in Jerusalem, October 14-18, for DISC 2013!

• Irina Calciu, Dave Dice, Yossi Lev, Victor Luchangco, Virendra Marathe and Nir Shavit's paper NUMA-Aware Reader-Writer Locks is accepted to PPoPP 2013.

• Yehuda Afek and Adam Morrison's paper Fast Concurrent Queues for x86 Processors is accepted to PPoPP 2013.

• Hillel Avni and Trevor Brown's paper Range Queries in Non-blocking k-ary Search Trees is accepted to OPODIS 2012.

• Yehuda Afek, Alexander Matveev and Nir Shavit's paper Pessimistic Software Lock Ellision is accepted to DISC 2012.

• Yehuda Afek, Haim Kaplan, Boris Korenfeld, Adam Morrison and Robert E. Tarjan's paper CBTree: A Practical Concurrent Self-Adjusting Search Tree is accepted to DISC 2012 and awarded best student paper.

• Alexander Matveev and Nir Shavit's paper Towards a Fully Pessimistic STM Model is accepted to TRANSACT 2012.

• Dave Dice, Virendra Marathe and Nir Shavit's paper Lock Cohorting: A General Technique for Designing NUMA Locks is accepted to PPoPP 2012.

• Yehuda Afek, Hillel Avni and Nir Shavit's paper Towards Consistency Oblivious Programming is accepted to OPODIS 2011 .

• Our group is awarded an ISF grant for Scalable and High Performance Software Execution on Chip Multiprocessors (2011-2014).

• Yehuda Afek, Michael Hakimi and Adam Morrison's paper Fast and Scalable Rendezvousing is accepted to DISC 2011 and awarded best student paper.

• Yehuda Afek, Yoav Cohen and Adam Morrison's paper Coping With Context Switches in Lock-Based Software Transactional Memory is accepted to SYSTOR 2011.

• Dave Dice, Virendra J. Marathe and Nir Shavit's paper Flat-combining NUMA locks is accepted to SPAA 2011.

• Yehuda Afek, Dave Dice and Adam Morrison's paper Cache Index-Aware Memory Allocation is accepted to ISMM 2011.

• Nir Shavit's review article Data Structures in the Multicore Age appears in the March issue of Communications of the ACM.

• Adam Morrison is awarded a 2011-2012 IBM PhD Fellowship.

• Hillel Avni, Yehuda Afek, Dave Dice and Nir Shavit's paper Efficient Lock Free Privatization is accepted to OPODIS 2010.

• Yehuda Afek, Guy Korland and Eitan Yanovsky's paper Quasi-Linearizability: Relaxed Consistency For Improved Concurrency is accepted to OPODIS 2010.

• Yehuda Afek, Guy Korland, Arie Zilberstein's paper Lowering STM Overhead with Static Analysis is accepted to LCPC 2010.