Cache aware optimization of stream programs

Sermulins, Janis; Thies, William; Rabbah, Rodric; Amarasinghe, Saman

doi:10.1145/1065910.1065927

Cited by 56 publications

(50 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Heuristic cache-aware scheduling of streaming programs on both single processors and multiprocessors has been studied by several research groups [15,21,25], but the proposed heuristics are all evaluated empirically. To the best of our knowledge, no previous work provides any theoretical guarantees comparable to ours.…”

Section: Related Workmentioning

confidence: 99%

“…Much of the research has focused on maximizing throughput, defined as the number of data items processed per unit time, and on minimizing delay or latency (the period between the time an input data item enters the computation and the time it affects an output data item) [6,13]. More recently, additional metrics have been considered, such as power consumption [29], memory usage [4,23], and cache efficiency [15,21,25].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cache-conscious scheduling of streaming applications

Agrawal

Fineman

Krage

et al. 2012

Proceedings of the Twenty-Fourth Annual ACM Symposium on Parallelism in Algorithms and Architectures

View full text Add to dashboard Cite

This paper considers the problem of scheduling streaming applications on uniprocessors in order to minimize the number of cache-misses. Streaming applications are represented as a directed graph (or multigraph), where nodes are computation modules and edges are channels. When a module fires, it consumes some data-items from its input channels and produces some items on its output channels. In addition, each module may have some state (either code or data) which represents the memory locations that must be loaded into cache in order to execute the module. We consider synchronous dataflow graphs where the input and output rates of modules are known in advance and do not change during execution. We also assume that the state size of modules is known in advance.Our main contribution is to show that for a large and important class of streaming computations, cache-efficient scheduling is essentially equivalent to solving a constrained graph partitioning problem. A streaming computation from this class has a cache-efficient schedule if and only if its graph has a low-bandwidth partition of the modules into components (subgraphs) whose total state fits within the cache, where the bandwidth of the partition is the number of data items that cross intercomponent channels per data item that enters the graph.Given a good partition, we describe a runtime strategy for scheduling two classes of streaming graphs: pipelines, where the graph consists of a single directed chain, and a fairly general class of directed acyclic graphs (dags) with some additional restrictions. The runtime scheduling strategy consists of adding large external buffers at the input and output edges of each component, allowing each component to be executed many times. Partitioning enables a reduction in cache misses in two ways. First, any items that are generated on edges internal to subgraphs are never written Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. SPAA '12, June 25-27, 2012, Pittsburgh, Pennsylvania, USA. Copyright 2012 ACM 978-1-4503-1213 out to memory, but remain in cache. Second, each subgraph is executed many times, allowing the state to be reused.We prove the optimality of this runtime scheduling for all pipelines and for dags that meet certain conditions on buffersize requirements. Specifically, we show that with constantfactor memory augmentation, partitioning on these graphs guarantees the optimal number of cache misses to within a constant factor. For the pipeline case, we also prove that such a partition can be found in polynomial time. For the dags we prove optimality if a good partition is provided; the partitioning problem itself is NP-complete.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Cache-conscious scheduling of streaming applications

Agrawal

Fineman

Krage

et al. 2012

Proceedings of the Twenty-Fourth Annual ACM Symposium on Parallelism in Algorithms and Architectures

View full text Add to dashboard Cite

show abstract

“…StreamIt [1] is both a stream language and an optimizing compiler for the RAW [4] and SMP [8] architectures. It is based on the cyclo-static dataflow model [6].…”

Section: Related Workmentioning

confidence: 99%

A Multidimensional Array Slicing DSL for Stream Programming

Castro

Louise

Barthou

2010

2010 International Conference on Complex, Intelligent and Software Intensive Systems

View full text Add to dashboard Cite

show abstract

“…This locality is typically achieved by restructuring the program code based on affinity analysis [8,10,12,13,14].…”

Section: Introductionmentioning

confidence: 99%

Instruction Hints for Super Efficient Data Caches

Tao

Hillenbrand

Marten

2009

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Data cache is a commodity in modern microprocessor systems. It is a fact that the size of data caches keeps growing up, however, the increase in application size goes faster. As a result, it is usually not possible to store the complete working set in the cache memory.This paper proposes an approach that allows the data access of some load/store instructions to bypass the cache memory. In this case, the cache space can be reserved for storing more frequently reused data. We implemented an analysis algorithm to identify the specific instructions and a simulator to model the novel cache architecture. The approach was verified using applications from MediaBench/MiBench benchmark suite and for all except one application we achieved huge gains in performance.

show abstract

Cache aware optimization of stream programs

Cited by 56 publications

References 14 publications

Cache-conscious scheduling of streaming applications

Cache-conscious scheduling of streaming applications

A Multidimensional Array Slicing DSL for Stream Programming

Instruction Hints for Super Efficient Data Caches

Contact Info

Product

Resources

About