Symmetries in data parallelism
نویسندگان
چکیده
منابع مشابه
Symmetries in Data Parallelism
A comprehensive formalisation of data-parallel symmetries in an imperative language paradigm without nesting is presented, which includes translational, a ne and access symmetries. A subtyping system which takes these symmetries into account is discussed. A multi-component type is introduced, with type inclusion in each component re ecting a diminished symmetry of a supertype compared to any of...
متن کاملData Parallelism in Java
Java supports threads and remote method invocation but it does neither support data parallel nor distributed programming. This paper discusses Java's shortcomings with respect to data parallel programming. It then presents countermeasures that allow for data parallel programming in Java. The technical contributions of this paper are twofold: a source-to-source transformation is presented that m...
متن کاملRelating Data-Parallelism and (And-) Parallelism in Logic Programs
Much work has been done in the áreas of and-parallelism and data parallelism in Logic Programs. Such work has proceeded to a certain extent in an independent fashion. Both types of parallelism offer advantages and disadvantages. Traditional (and-) parallel models offer generality, being able to exploit parallelism in a large class of programs (including that exploited by data parallelism techni...
متن کاملClustered Data Parallelism
Many data layout optimizations cluster data accesses and memory into high-locality groups in order to optimize for the memory hierarchy. In this paper, we demonstrate that similar clustering program transformations enable efficient vectorization. We call this approach clustered data parallelism (CDP). CDP enables fast and power-efficient parallelism by partitioning a data structure into cluster...
متن کاملConstrained Data-Driven Parallelism
In data-driven parallelism, changes to data spawn new tasks, which may change more data, spawning yet more tasks. Computation propagates until no further changes occur. Benefits include increasing opportunities for finegrained parallelism, avoiding redundant work, and supporting incremental computations on large data sets. Nonetheless, data-driven parallelism can be problematic. For example, co...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: The Computer Journal
سال: 1995
ISSN: 0010-4620,1460-2067
DOI: 10.1093/comjnl/38.5.365