Learning Forward Reuse Distance

07/31/2020
by   Pengcheng Li, et al.
0

Caching techniques are widely used in the era of cloud computing from applications, such as Web caches to infrastructures, Memcached and memory caches in computer architectures. Prediction of cached data can greatly help improve cache management and performance. The recent advancement of deep learning techniques enables the design of novel intelligent cache replacement policies. In this work, we propose a learning-aided approach to predict future data accesses. We find that a powerful LSTM-based recurrent neural network model can provide high prediction accuracy based on only a cache trace as input. The high accuracy results from a carefully crafted locality-driven feature design. Inspired by the high prediction accuracy, we propose a pseudo OPT policy and evaluate it upon 13 real-world storage workloads from Microsoft Research. Results demonstrate that the new cache policy improves state-of-art practical policies by up to 19.2 OPT on average.

READ FULL TEXT

page 6

page 8

page 10

research
05/02/2018

ECI-Cache: A High-Endurance and Cost-Efficient I/O Caching Scheme for Virtualized Platforms

In recent years, high interest in using Virtual Machines (VMs) in data c...
research
05/31/2019

Evaluating Memento Service Optimizations

Services and applications based on the Memento Aggregator can suffer fro...
research
11/13/2020

Phoebe: Reuse-Aware Online Caching with Reinforcement Learning for Emerging Storage Models

With data durability, high access speed, low power efficiency and byte a...
research
06/15/2020

Addressing Variability in Reuse Prediction for Last-Level Caches

Last-Level Cache (LLC) represents the bulk of a modern CPU processor's t...
research
09/30/2019

Optimizing GPU Cache Policies for MI Workloads

In recent years, machine intelligence (MI) applications have emerged as ...
research
04/06/2022

SqueezeNeRF: Further factorized FastNeRF for memory-efficient inference

Neural Radiance Fields (NeRF) has emerged as the state-of-the-art method...
research
08/15/2018

Making Belady-Inspired Replacement Policies More Effective Using Expected Hit Count

Memory-intensive workloads operate on massive amounts of data that canno...

Please sign up or login with your details

Forgot password? Click here to reset