Homunculus : Auto-Generating Efficient Data-Plane ML Pipelines for Datacenter Networks

↓ Direkt till sidans innehåll
↓ Direkt till sidans sekundära innehåll (sidomenyn)

Search: id:"swepub:oai:lup.lub.lu.se:e764ee2d-5cf5-40b2-84da-d507a0c8a070" > Homunculus : Auto-G...

1 of 1
Previous record
Next record
To hitlist

Details
MARC

Swamy, TusharStanford University (author)

Homunculus : Auto-Generating Efficient Data-Plane ML Pipelines for Datacenter Networks

Article/chapterEnglish2023

Publisher, publication year, extent ...

2023-03-25
New York, NY, USA :ACM,2023
14 s.

Numbers

LIBRIS-ID:oai:lup.lub.lu.se:e764ee2d-5cf5-40b2-84da-d507a0c8a070
https://lup.lub.lu.se/record/e764ee2d-5cf5-40b2-84da-d507a0c8a070URI
https://doi.org/10.1145/3582016.3582022DOI

Supplementary language notes

Language:English
Summary in:English

Part of subdatabase

SwePubSwePub

Classification

Subject category:kon swepub-publicationtype
Subject category:ref swepub-contenttype

Notes

Support for Machine Learning (ML) applications in networking has significantly improved over the last decade. The availability of public datasets and programmable switching fabrics (including low-level languages to program them) presents a full-stack to the programmer for deploying in-network ML. However, the diversity of tools involved, coupled with complex optimization tasks of ML model design and hyperparameter tuning while complying with the network constraints (like throughput and latency), puts the onus on the network operator to be an expert in ML, network design, and programmable hardware. We present Homunculus, a high-level framework that enables network operators to specify their ML requirements in a declarative rather than imperative way. Homunculus takes as input the training data and accompanying network and hardware constraints, and automatically generates and installs a suitable model onto the underlying switching target. It performs model design-space exploration, training, and platform code-generation as compiler stages, leaving network operators to focus on acquiring high-quality network data. Our evaluations on real-world ML applications show that Homunculus's generated models achieve up to 12% better F1 scores compared to hand-tuned alternatives, while operating within the resource limits of the underlying targets. We further demonstrate the high performance and increased reactivity (seconds to nanoseconds) of the generated models on emerging per-packet ML platforms to showcase Homunculus's timely and practical significance.

Subject headings and genre

TEKNIK OCH TEKNOLOGIER Elektroteknik och elektronik Datorsystem hsv//swe
ENGINEERING AND TECHNOLOGY Electrical Engineering, Electronic Engineering, Information Engineering Computer Systems hsv//eng
ML Compilers
Per-packet ML
Self-driving Networks

Added entries (persons, corporate bodies, meetings, titles ...)

Zulfiqar, AnnusPurdue University (author)
Nardi, LuigiLund University,Lunds universitet,Robotik och Semantiska System,Institutionen för datavetenskap,Institutioner vid LTH,Lunds Tekniska Högskola,LTH profilområde: AI och digitalisering,LTH profilområden,Robotics and Semantic Systems,Department of Computer Science,Departments at LTH,Faculty of Engineering, LTH,LTH Profile Area: AI and Digitalization,LTH Profile areas,Faculty of Engineering, LTH,Stanford University(Swepub:lu)lu6078na (author)
Shahbaz, MuhammadPurdue University (author)
Olukotun, KunleStanford University (author)
Aamodt, Tor M. (editor)
Jerger, Natalie Enright (editor)
Swift, Michael (editor)
Stanford UniversityPurdue University (creator_code:org_t)

Related titles

In:ASPLOS 2023 - Proceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating SystemsNew York, NY, USA : ACM3, s. 329-3429781450399180

Internet link

Find in a library

ASPLOS 2023 - Proceedings of the 28th ACM International Conference on Architectu... (Search for host publication in LIBRIS)

To the university's database

1 of 1
Previous record
Next record
To hitlist

Find more in SwePub

By the author/editor: Swamy, Tushar; Zulfiqar, Annus; Nardi, Luigi; Shahbaz, Muhamma ...; Olukotun, Kunle; Aamodt, Tor M.; show more...; Jerger, Natalie ...; Swift, Michael; show less...

About the subject

ENGINEERING AND TECHNOLOGY: ENGINEERING AND ...; and Electrical Engin ...; and Computer Systems

Articles in the publication: ASPLOS 2023 - Pr ...

By the university: Lund University

Search outside SwePub

Extend your search to:: Google; Google Book Search; Google Scholar

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.

LIBRIS.kb.se