2021/08/02 by Jennifer Switzer, Barath Raghavan, Switzer, Jennifer +1 · 1 citation
Computer Science · #Advanced Data Storage Technologies #Caching and Content Delivery #Cloud Computing and Resource Management #Distributed #FOS: Computer and information sciences #Parallel #and Cluster Computing (cs.DC)
paper · pdf · doi:10.48550/arxiv.2108.01035
openalex publication_date 2021/08/02 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Coping with the intermittency of renewables is a fundamental challenge, with\nload shifting and grid-scale storage as key responses. We propose Information\nBatteries (IB), in which energy is stored in the form of information --\nspecifically, the results of completed computational tasks. Information\nBatteries thus provide storage through speculative load shifting, anticipating\ncomputation that will be performed in the future.\n We take a distributed systems perspective, and evaluate the extent to which\nan IB storage system can be made practical through augmentation of compiler\ntoolchains, key-value stores, and other important elements in modern\nhyper-scale compute. In particular, we implement one specific IB prototype by\naugmenting the Rust compiler to enable transparent function-level\nprecomputation and caching. We evaluate the overheads this imposes, along with\nmacro-level job prediction and power prediction. We also evaluate the space of\noperation for an IB system, to identify the best case efficiency of any IB\nsystem for a given power and compute regime.\n