It's Time to Think About an Operating System for Near Data Processing Architectures

Barbalace, Antonio; Iliopoulos, Anthony; Rauchfuss, Holm; Brasche, Goetz

doi:10.1145/3102980.3102990

Cited by 24 publications

(15 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Like this prior work, AMS focuses on how to schedule threads across an asymmetric system to maximize system-wide performance. Unlike this prior work, AMS aims to schedule threads with no program modifications and transparently to users, similar to how OS-level schedulers manage symmetric systems, as recent work on OS for NDP systems advocates [7].…”

Section: A Pim and Ndp Systemsmentioning

confidence: 99%

Adaptive Scheduling for Systems with Asymmetric Memory Hierarchies

Tsai

Chen

Sánchez

2018

2018 51st Annual IEEE/ACM International Symposium on Microarchitecture (MICRO)

View full text Add to dashboard Cite

Conventional multicores rely on deep cache hierarchies to reduce data movement. Recent advances in die stacking have enabled near-data processing (NDP) systems that reduce data movement by placing cores close to memory. NDP cores enjoy cheaper memory accesses and are more area-constrained, so they use shallow cache hierarchies instead. Since neither shallow nor deep hierarchies work well for all applications, prior work has proposed systems that incorporate both. These asymmetric memory hierarchies can be highly beneficial, but they require scheduling computation to the right hierarchy. We present AMS, an adaptive scheduler that automatically finds high-quality thread-to-hierarchy mappings. AMS monitors threads, accurately models their performance under different hierarchies and core types, and adapts algorithms first proposed for cache partitioning to produce high-quality schedules. AMS is cheap enough to use online, so it adapts to program phases, and performs within 1% of an exhaustive-search scheduler. As a result, AMS outperforms asymmetry-oblivious schedulers by up to 37% and by 18% on average.

show abstract

Section: A Pim and Ndp Systemsmentioning

confidence: 99%

Adaptive Scheduling for Systems with Asymmetric Memory Hierarchies

Tsai

Chen

Sánchez

2018

2018 51st Annual IEEE/ACM International Symposium on Microarchitecture (MICRO)

View full text Add to dashboard Cite

show abstract

“…A popular technique to avoid moving large amounts of data over the network and avoid serialization and de-serialization overhead is to push data preprocessing routines close to where the data is stored, especially when considering modern storage equipped with processing capability. Also known as near-data processing [30], this feature would greatly benefit stream processing yet current ingestion/storage solutions do not offer native support for it.…”

Section: Missing Featuresmentioning

confidence: 99%

Towards a unified storage and ingestion architecture for stream processing

Marcu

Costan

Antoniu

et al. 2017

2017 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

Abstract-Big Data applications are rapidly moving from a batch-oriented execution model to a streaming execution model in order to extract value from the data in real-time. However, processing live data alone is often not enough: in many cases, such applications need to combine the live data with previously archived data to increase the quality of the extracted insights. Current streaming-oriented runtimes and middlewares are not flexible enough to deal with this trend, as they address ingestion (collection and pre-processing of data streams) and persistent storage (archival of intermediate results) using separate services. This separation often leads to I/O redundancy (e.g., write data twice to disk or transfer data twice over the network) and interference (e.g., I/O bottlenecks when collecting data streams and writing archival data simultaneously). In this position paper, we argue for a unified ingestion and storage architecture for streaming data that addresses the aforementioned challenge. We identify a set of constraints and benefits for such a unified model, while highlighting the important architectural aspects required to implement it in real life. Based on these aspects, we briefly sketch our plan for future work that develops the position defended in this paper.

show abstract

“…Instead, previous works introduced (mostly use-case-specific) point solutions, which focus primarily on performance, but hinder programmability, portability, and wider adoption of the technology. Thus, it is urgent to introduce a programming model and framework, as well as associated hardware and software protocols and interfaces, to enable users to efficiently tap into the compute resources available on emerging SSDs [12], independently of how the data was originally stored (i.e., file system neutral).…”

Section: Introductionmentioning

confidence: 99%