An Overview of Cache Optimization Techniques and Cache-Aware Numerical Algorithms

Kowarschik, Markus; Weiß, Christian

doi:10.1007/3-540-36574-5_10

Cited by 108 publications

(100 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Many techniques change the order in which data is accessed to increase the effectiveness of processor caches by altering software at the source code level. These optimizations, known as data access optimizations [13], focus on changing the manner in which loops are executed. One technique, known as loop interchange [30], can be used to reorder multiple loops to maximize the data access of common elements in respect to time, referred to as temporal locality [2,31,30,24].…”

Section: Related Workmentioning

confidence: 99%

“…Loop fusion maximizes temporal locality by merging multiple loops into a single loop and altering data access order [25,12,6]. Yet another technique for improving software cache effectiveness is to utilize prefetch instructions, which retrieves data from memory into the cache before the data is requested by the application [13]. Prefetch instructions inserted manually into software at the source code level can significantly reduce memory latency and/or cache miss rate [7,9].…”

Section: Related Workmentioning

confidence: 99%

“…Likewise, Manjikian et al [16] demonstrated a 25% reduction in execution time as a result of modifying the source-code of the executing software to use cache partitioning. Many optimization techniques [21,17,27] increase cache hit rate by enhancing source code to increase temporal locality of data accesses, which defines the proximity with which shared data is accessed in terms of time [13]. For example, loop interchange and loop fusion techniques can increase temporal locality of accessed data by modifying application source code to change the order in which application data is written to and read from a processor cache [13,16].…”

Section: Introductionmentioning

confidence: 99%

“…Many optimization techniques [21,17,27] increase cache hit rate by enhancing source code to increase temporal locality of data accesses, which defines the proximity with which shared data is accessed in terms of time [13]. For example, loop interchange and loop fusion techniques can increase temporal locality of accessed data by modifying application source code to change the order in which application data is written to and read from a processor cache [13,16]. Increasing temporal locality increases the probability that data common to multiple tasks persists in the cache, thereby reducing cache-misses and software execution time [13,16].…”

Section: Introductionmentioning

confidence: 99%

“…For example, loop interchange and loop fusion techniques can increase temporal locality of accessed data by modifying application source code to change the order in which application data is written to and read from a processor cache [13,16]. Increasing temporal locality increases the probability that data common to multiple tasks persists in the cache, thereby reducing cache-misses and software execution time [13,16].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Optimizing Integrated Application Performance with Cache-Aware Metascheduling

Dougherty

White

Kegley

et al. 2011

On the Move to Meaningful Internet Systems: OTM 2011

View full text Add to dashboard Cite

Abstract. Integrated applications running in multi-tenant environments are often subject to quality-of-service (QoS) requirements, such as resource and performance constraints. It is hard to allocate resources between multiple users accessing these types of applications while meeting all QoS constraints, such as ensuring users complete execution prior to deadlines. Although a processor cache can reduce the time required for the tasks of a user to execute, multiple task execution schedules may exist that meet deadlines but differ in cache utilization efficiency. Determining which task execution schedules will utilize the processor cache most efficiently and provide the greatest reductions in execution time is hard without jeopardizing deadlines. The work in this paper provides three key contributions to increasing the execution efficiency of integrated applications in multi-tenant environments while meeting QoS constraints. First, we present cache-aware metascheduling, which is a novel approach to modifying system execution schedules to increase cache-hit rate and reduce system execution time. Second, we apply cache-aware metascheduling to 11 simulated software systems to create 2 different execution schedules per system. Third, we empirically evaluate the impact of using cache-aware metascheduling to alter task schedules to reduce system execution time. Our results show that cache-aware metascheduling increases cache performance, reduces execution time, and satisfies scheduling constraints and safety requirements without requiring significant hardware or software changes.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Optimizing Integrated Application Performance with Cache-Aware Metascheduling

Dougherty

White

Kegley

et al. 2011

On the Move to Meaningful Internet Systems: OTM 2011

View full text Add to dashboard Cite

show abstract

On parallel computation of indicators of centrality and its application to speeding up RDA

Sierra

Carriegos

2019

Comp and Math Methods

View full text Add to dashboard Cite

Random decentralized algorithm (RDA) dynamically extends networks, hiding their central structure so that they becomes less vulnerable to attacks. To assess the efficacy of RDA, we have to monitor the values for several indicators of centrality, which must follow a decremental pattern as the algorithm progresses. Although optimized algorithms exist for computing of most of these indicators, they are still time consuming and are even infeasible to apply to big enough graphs like the ones representing social networks or extensive enough computer networks, which limit the scope of application of RDA to small networks. In this paper, we present a threading‐based parallel implementation in C language of some optimal algorithms for computing the indicators of centrality used by RDA. We have tested our software in several platforms, including the Supercomputer Calendula, and our parallel version greatly reduces the execution time of their sequential (nonparallel) counterpart. Our application is multiplatform and portable, working on any machine with several logical processors, which is capable of compiling and running C language code. The speedup of our solution allows us to apply RDA algorithm to networks with thousand of nodes.

show abstract

A cache‐oblivious self‐adaptive full multigrid method

Mehl

Weinzierl

Zenger

2006

Numerical Linear Algebra App

View full text Add to dashboard Cite

SUMMARYThis paper presents a new e cient way to implement multigrid algorithms on adaptively reÿned grids. To cope with todays demands in high-performance computing, we cannot do without such highly sophisticated numerical methods. But if we do not implement them very carefully, we lose a lot of e ciency in terms of memory usage: using trees for the storage of hierarchical multilevel data causes a large amount of non-local (in terms of the physical memory space) data accesses, and often requires the storage of pointers to neighbours to allow the evaluation of discrete operators (di erence stencils, restrictions, interpolations, etc.). The importance of this problem becomes clear if we remember that storage and not the CPUs is the bottleneck on modern computers.We established a cache-oblivious and storage-minimizing algorithm based on the concept of spacetree grids combined with a cell-oriented operator evaluation, a linear ordering of grid cells along a space-ÿlling curve, and a sophisticated construction of linearly processed data structures for vertex data. In this context, we could show that the implementation of a dynamically adaptive F-cycle is, ÿrst, very natural and, second, does not cause any overhead in terms of storage usage and access as adaptivity and multilevel data do not disturb the linear processing order of our data structures.

show abstract

An Overview of Cache Optimization Techniques and Cache-Aware Numerical Algorithms

Cited by 108 publications

References 0 publications

Optimizing Integrated Application Performance with Cache-Aware Metascheduling

Optimizing Integrated Application Performance with Cache-Aware Metascheduling

On parallel computation of indicators of centrality and its application to speeding up RDA

A cache‐oblivious self‐adaptive full multigrid method

Contact Info

Product

Resources

About