A
Aleksandar Dragojevic
Researcher at Microsoft
Publications - 32
Citations - 2043
Aleksandar Dragojevic is an academic researcher from Microsoft. The author has contributed to research in topics: Transactional memory & Software transactional memory. The author has an hindex of 19, co-authored 31 publications receiving 1741 citations. Previous affiliations of Aleksandar Dragojevic include École Normale Supérieure & École Polytechnique Fédérale de Lausanne.
Papers
More filters
Proceedings ArticleDOI
FaRM: fast remote memory
TL;DR: The design and implementation of FaRM is described, a new main memory distributed computing platform that exploits RDMA to improve both latency and throughput by an order of magnitude relative to state of the art main memory systems that use TCP/IP.
Proceedings ArticleDOI
No compromises: distributed transactions with consistency, availability, and performance
Aleksandar Dragojevic,Dushyanth Narayanan,Edmund B. Nightingale,Matthew Renzelmann,Alex Shamis,Anirudh Badam,Miguel Castro +6 more
TL;DR: It is shown that a main memory distributed computing platform called FaRM can provide distributed transactions with strict serializability, high performance, durability, and high availability in modern data centers.
Proceedings ArticleDOI
Stretching transactional memory
TL;DR: SwissTM is lock- and word-based and uses a new two-phase contention manager that ensures the progress of long transactions while inducing no overhead on short ones, and outperforms state-of-the-art STM implementations, namely RSTM, TL2, and TinySTM.
Journal ArticleDOI
Why STM can be more than a research toy
TL;DR: Despite earlier claims, Software Transactional Memory outperforms sequential code in terms of memory efficiency.
Proceedings ArticleDOI
Preventing versus curing: avoiding conflicts in transactional memories
TL;DR: Shrink is presented, a scheduler that bases its prediction of transactional accesses on the access patterns of the past transactions from the same thread, and uses a novel heuristic, which is called serialization affinity, to schedule transactions with a probability proportional to the current amount of contention.