Junkyard Computing: Repurposing Discarded Smartphones to Minimize Carbon
Status:: 🟩
Links::
Metadata
Authors:: Switzer, Jennifer; Marcano, Gabriel; Kastner, Ryan; Pannuto, Pat
Title:: Junkyard Computing: Repurposing Discarded Smartphones to Minimize Carbon
Date:: 2023
Publisher:: Association for Computing Machinery
URL:: https://dl.acm.org/doi/10.1145/3575693.3575710
DOI:: 10.1145/3575693.3575710
Bibliography
Switzer J, Marcano G, Kastner R, Pannuto P (2023) Junkyard Computing: Repurposing Discarded Smartphones to Minimize Carbon. In: Proceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2. Association for Computing Machinery, New York, NY, USA, pp 400–412
Zotero
Type:: #zotero/conferencePaper
Keywords:: []
Relations
Abstract
1.5 billion smartphones are sold annually, and most are decommissioned less than two years later. Most of these unwanted smartphones are neither discarded nor recycled but languish in junk drawers and storage units. This computational stockpile represents a substantial wasted potential: modern smartphones have increasingly high-performance and energy-efficient processors, extensive networking capabilities, and a reliable built-in power supply. This project studies the ability to reuse smartphones as "junkyard computers." Junkyard computers grow global computing capacity by extending device lifetimes, which supplants the manufacture of new devices. We show that the capabilities of even decade-old smartphones are within those demanded by modern cloud microservices and discuss how to combine phones to perform increasingly complex tasks. We describe how current operation-focused metrics do not capture the actual carbon costs of compute. We propose Computational Carbon Intensity---a performance metric that balances the continued service of older devices with the superlinear runtime improvements of newer machines. We use this metric to redefine device service lifetime in terms of carbon efficiency. We develop a cloudlet of reused Pixel 3A phones. We analyze the carbon benefits of deploying large, end-to-end microservice-based applications on these smartphones. Finally, we describe system architectures and associated challenges to scale to cloudlets with hundreds and thousands of smartphones.
Notes & Annotations
📑 Annotations (imported on 2023-05-26#14:50:06)
Computational Carbon Intensity (CCI) measures the lifetime carbon impact of a device versus the lifetime useful compute it performs.
We apply CCI to old servers, old laptops, and old smartphones. While each device type shows potential as carbon-saving hardware, we find that used smartphones (repurposed as general-purpose compute nodes) offer the best potential for carbon impact.
Life-cycle assessment (LCA) seeks to characterize the lifetime environmental impact of a product across multiple metrics, including carbon. LCAs have been performed for smartphones, laptops, and servers [15, 17]. Manufacturing accounts for 70 − 90% of the lifetime carbon footprint for phones and laptops. For high-performance computing devices, manufacturing accounts for 20 − 50% of the lifetime carbon cost [15, 17, 31].
LCAs measure the total carbon footprint of a device across its lifetime; our approach (CCI) is different in that we consider the amortized carbon footprint of each unit of computational work.
Smart charging opportunistically charges the devices whenever the grid-level carbon intensity falls below a tunable threshold. This threshold is based on historic grid characteristics and the charged device.
The major takeaways of this work are: (1) For specific workloads, clusters of repurposed phones are cheaper and more carbon efficient than traditional servers. (2) More broadly, scavenging unwanted equipment shows excellent potential for building economic and carbon-efficient systems, especially when renewable energy is plentiful. (3) Sustainability has operational and manufacturing facets; manufacturing dominates as operating trends towards zero with cleaner energy mixes. (4) Accurate LCA information is essential for carbon-based analyses; it would be beneficial if more ICT manufacturers published this information, including cloud providers who build custom systems.