Reuse-based online models for caches

Sen, Rathijit; Wood, David А.

doi:10.1145/2465529.2465756

Cited by 38 publications

(12 citation statements)

References 46 publications

(23 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In [50], an analysis over memory workloads shows that the stack distance of memory accesses follows a Zipf distribution law, which is also observed in other application domains such as web caching and disk I/O [51], [52]. A low-overhead hardware mechanism for calculating the stack distance is proposed in [53]. The accuracy of modelling LRUbased caching techniques is discussed in [54].…”

Section: Previous Studiesmentioning

confidence: 99%

An Analytical Model for Performance and Lifetime Estimation of Hybrid DRAM-NVM Main Memories

Salkhordeh

Mutlu

Asadi

2019

IEEE Trans. Comput.

View full text Add to dashboard Cite

Emerging Non-Volatile Memories (NVMs) have promising advantages (e.g., lower idle power, higher density, and nonvolatility) over the existing predominant main memory technology, DRAM. Yet, NVMs also have disadvantages (e.g., longer latencies, higher active power, and limited endurance). System architects are therefore examining hybrid DRAM-NVM main memories to enable the advantages of NVMs while avoiding the disadvantages as much as possible. Unfortunately, the hybrid memory design space is very large and complex due to the existence of very different types of NVMs and their rapidly-changing characteristics. Therefore, optimization of performance and lifetime of hybrid memory based computing platforms and their experimental evaluation using traditional simulation methods can be very time-consuming and sometimes even impractical. As such, it is necessary to develop a fast and flexible analytical model to estimate the performance and lifetime of hybrid memories on various workloads. This paper presents an analytical model for hybrid memories based on Markov decision processes. The proposed model estimates the hit ratio and lifetime for various configurations of DRAM-NVM hybrid main memories. Our model also provides accurate estimation of the effect of data migration policies on the hybrid memory hit ratio (i.e., percentage of accesses supplied by either DRAM or NVM), one of the most important factors in hybrid memory performance and lifetime. Such an analytical model can aid designers to tune hybrid memory configurations to improve performance and/or lifetime. We present several optimizations that make our model more efficient while maintaining its accuracy. Our experimental evaluations conducted using the PARSEC benchmark suite show that the proposed model (a) accurately predicts the hybrid memory hit ratio compared to the state-of-the-art hybrid memory simulators with an average (maximum) error of 4.61% (13.6%) on a commodity server (equipped with 192GB main memory and quad-core Xeon processor), (b) accurately estimates the NVM lifetime with an average (maximum) error of 2.93% (8.8%), and (c) is on average (up to) 4x (10x) faster than conventional state-of-the-art simulation platforms for hybrid memories.

show abstract

Section: Previous Studiesmentioning

confidence: 99%

An Analytical Model for Performance and Lifetime Estimation of Hybrid DRAM-NVM Main Memories

Salkhordeh

Mutlu

Asadi

2019

IEEE Trans. Comput.

View full text Add to dashboard Cite

show abstract

“…First, we consider a single set count s. Following the idea of per-set reuse distance [16,18], we measure the partial footprint for accesses to each cache set S i as its mapped footprint, which is pfp(S i , x). It is measured by first decomposing the execution trace into TP traces, one for each S i .…”

Section: Mapping the Footprintmentioning

confidence: 99%

“…Sen and Wood [18] and Nugteren et al [16] created the first accurate solution, which measures the reuse distance in each cache set and then uses the per-set reuse distance to predict the miss ratio. Sen and Wood called it per-set locality [18]. Following Sen and Wood, we call this model per set reuse distance, or Set-RD in short.…”

Section: Three Models Of Associative Cachementioning

confidence: 99%

See 1 more Smart Citation

Footprint modeling of cache associativity and granularity

Luo

Chen

Liu

et al. 2018

Proceedings of the International Symposium on Memory Systems

View full text Add to dashboard Cite

Two common techniques in efficient caching are associativity and sub-block granularity. This short paper presents a parameterized and composable model for each of the two techniques. It shows how the new models are more general, accurate or efficient than previous modeling solutions in either technique, and how they can be used together to model the cache implemented with both techniques, i.e. sub-block set associative cache. CCS Concepts • Computing methodologies → Modeling methodologies;

show abstract

“…There are different types of techniques to analyze LLC behavior, such as analytical model [1], [2], trace-driven simulation [3]- [5] and profiling-based performance prediction [6]- [8]. Analytical models typically provide insights on program behavior for performance prediction.…”

Section: Introductionmentioning

confidence: 99%

A Gaussian Set Sampling Model for Efficient Shared Cache Profiling on Multi-Cores

Zhang

Ling

et al. 2019

IEEE Access

View full text Add to dashboard Cite

The last level cache (LLC) has significant impact to system performance on modern multi-core processors. But as cache sizes reach several megabytes and more, the overhead of exploring performance on LLC greatly increases as well. To improve the efficiency of performance analysis, we propose a setsampling-based cache profiling model for the performance analysis on multi-core LLC. We first explore the memory access distributions on LLC by developing a low-overhead stress-application-based method. The results show that memory access distributions can be approximated by Gaussian distribution function. Based on this observation, a Gaussian-distribution-based set sampling model is proposed which can predict program performance with limited representative samples. We evaluate our model on a contemporary multicore machine and show that 1) the proposed method can precisely predict program performance on LLC under different contention intensities and 2) our method can achieve similar precision with less samples compared to widely adopted set sampling methods such as the random sampling and the continuous address sampling.

show abstract

Reuse-based online models for caches

Cited by 38 publications

References 46 publications

An Analytical Model for Performance and Lifetime Estimation of Hybrid DRAM-NVM Main Memories

An Analytical Model for Performance and Lifetime Estimation of Hybrid DRAM-NVM Main Memories

Footprint modeling of cache associativity and granularity

A Gaussian Set Sampling Model for Efficient Shared Cache Profiling on Multi-Cores

Contact Info

Product

Resources

About