Proceedings of the 15th ACM/SPEC International Conference

Years of planning have gone into preparing for ICPE 2024 in London, UK. For the first time in UK, the organization of ICPE has generated a great deal of excitement and expectation of productive interactions between the usual participants of ICPE conferences, the members of the various SPEC working groups, and a desire to increase the involvement of the local scientific community with ICPE.

It is our pleasure to welcome you to the 15th ACM/SPEC International Conference on Performance Engineering (ICPE), hosted at South Kensington, London, UK, from May 7-11, 2024. ICPE is the leading international forum for presenting and discussing novel ideas, innovations, trends and experiences in the field of performance engineering.

ICPE formed from merging the ACM Workshop on Software Performance (WOSP, since 1998) and the SPEC International Performance Engineering Workshop (SIPEW, since 2008). Despite the peculiar time we are all living in around the world, we are pleased to introduce an exciting program, which is the result of hard work by the authors, the program committee, and the conference organizers.

ICPE 2024 Welcome

General Chairs

Simonetta Balsamo, Ca' Foscari University of Venice, Italy
William Knottenbelt, Imperial College London, UK

Program Chairs

Cristina L. Abad, Escuela Superior Politecnica del Litoral, Ecuador
Weiyi Shang, University of Waterloo, Canada

ICPE 2024 Conference Organization

ICPE 2024 Sponsors & Supporters

Table of contents

Keynote Talk 1

How the Cloud made Performance Appear on the Board Agenda:

Authors
Manzoor Mohammed
DOI
10.1145/3629526.3645056
Pages
1 -- 1

Session 1a: Benchmarking

ShuffleBench: A Benchmark for Large-Scale Data Shuffling Operations with Distributed Stream Processing Frameworks:

Authors
Sören Henning
Adriano Vogel
Michael Leichtfried
Otmar Ertl
Rick Rabiser
DOI
10.1145/3629526.3645036
Pages
2 -- 13

Session 1b: Programming Langauge and Software

Vectorized Intrinsics Can Be Replaced with Pure Java Code without Impairing Steady-State Performance:

Authors
Júnior Löff
Filippo Schiavio
Andrea Rosà
Matteo Basso
Walter Binder
DOI
10.1145/3629526.3645051
Pages
14 -- 24

Rethinking 'Complement' Recommendations at Scale with SIMD:

Authors
Shrey Pandey
Saikat Kumar Das
Hrishikesh V. Ganu
Satyajeet Singh
DOI
10.1145/3629526.3645041
Pages
25 -- 36

An Adaptive Logging System (ALS): Enhancing Software Logging with Reinforcement Learning Techniques:

Authors
Amirmahdi Khosravi Tabrizi
Naser Ezzati-Jivan
Francois Tetreault
DOI
10.1145/3629526.3645033
Pages
37 -- 47

Time Series Forecasting of Runtime Software Metrics: An Empirical Study:

Authors
Federico Di Menna
Luca Traini
Vittorio Cortellessa
DOI
10.1145/3629526.3645049
Pages
48 -- 59

An Empirical Analysis of Common OCI Runtimes' Performance Isolation Capabilities:

Authors
Simon Volpert
Sascha Winkelhofer
Stefan Wesner
Jörg Domaschka
DOI
10.1145/3629526.3645044
Pages
60 -- 70

Session 2a: Hardware

An Experimental Setup to Evaluate RAPL Energy Counters for Heterogeneous Memory:

Authors
Lukas Alt
Anara Kozhokanova
Thomas Ilsche
Christian Terboven
Matthias S. Mueller
DOI
10.1145/3629526.3645052
Pages
71 -- 82

Using Evolutionary Algorithms to Find Cache-Friendly Generalized Morton Layouts for Arrays:

Authors
Stephen Nicholas Swatman
Ana-Lucia Varbanescu
Andy D. Pimentel
Andreas Salzburger
Attila Krasznahorkay
DOI
10.1145/3629526.3645034
Pages
83 -- 94

Energy Efficiency Features of the Intel Alder Lake Architecture:

Authors
Robert Schöne
Markus Velten
Daniel Hackenberg
Thomas Ilsche
DOI
10.1145/3629526.3645040
Pages
95 -- 106

Developing Index Structures in Persistent Memory Using Spot-on Optimizations with DRAM:

Authors
Xingsheng Zhao
Prajwal Challa
Chen Zhong
Song Jiang
DOI
10.1145/3629526.3645032
Pages
107 -- 117

Keynote Talk 2

What does Performance Mean for Large Language Models?:

Authors
Jane Hillston
DOI
10.1145/3629526.3649130
Pages
118 -- 118

Session 3a: Micro-services 1

InstantOps: A Joint Approach to System Failure Prediction and Root Cause Identification in Microserivces Cloud-Native Applications:

Authors
Raphael Rouf
Mohammadreza Rasolroveicy
Marin Litoiu
Seema Nagar
Prateeti Mohapatra
Pranjal Gupta
Ian Watts
DOI
10.1145/3629526.3645047
Pages
119 -- 129

Session 3b: Resource Management

Daedalus: Self-Adaptive Horizontal Autoscaling for Resource Efficiency of Distributed Stream Processing Systems:

Authors
Benjamin J. J. Pfister
Dominik Scheinert
Morgan K. Geldenhuys
Odej Kao
DOI
10.1145/3629526.3645042
Pages
130 -- 141

Demeter: Resource-Efficient Distributed Stream Processing under Dynamic Loads with Multi-Configuration Optimization:

Authors
Morgan K. Geldenhuys
Dominik Scheinert
Odej Kao
Lauritz Thamsen
DOI
10.1145/3629526.3645048
Pages
142 -- 153

BFQ, Multiqueue-Deadline, or Kyber? Performance Characterization of Linux Storage Schedulers in the NVMe Era:

Authors
Zebin Ren
Krijn Doekemeijer
Nick Tehrany
Animesh Trivedi
DOI
10.1145/3629526.3645053
Pages
154 -- 165

The Cost of Simplicity: Understanding Datacenter Scheduler Programming Abstractions:

Authors
Aratz Manterola Lasa
Sacheendra Talluri
Tiziano De Matteis
Alexandru Iosup
DOI
10.1145/3629526.3645038
Pages
166 -- 177

Session 4a: GPU and AI

Accelerating ML Workloads using GPU Tensor Cores: The Good, the Bad, and the Ugly:

Authors
Bagus Hanindhito
Lizy K. John
DOI
10.1145/3629526.3653835
Pages
178 -- 189

MalleTrain: Deep Neural Networks Training on Unfillable Supercomputer Nodes:

Authors
Xiaolong Ma
Feng Yan
Lei Yang
Ian Foster
Michael E. Papka
Zhengchun Liu
Rajkumar Kettimuthu
DOI
10.1145/3629526.3645035
Pages
190 -- 200

Leftovers for LLaMA:

Authors
Ravi Kumar Singh
Likhith Bandamudi
Shruti Kunde
Mayank Mishra
Rekha Singhal
DOI
10.1145/3629526.3645045
Pages
201 -- 210

Processing Natural Language on Embedded Devices: How Well Do Modern Models Perform?:

Authors
Souvika Sarkar
Mohammad Fakhruddin Babar
Md Mahadi Hassan
Monowar Hasan
Shubhra Kanti Karmaker Santu
DOI
10.1145/3629526.3645054
Pages
211 -- 222

Keynote Talk 3

Optimizing Edge AI: Performance Engineering in Resource-Constrained Environments:

Authors
Giuliano Casale
DOI
10.1145/3629526.3649131
Pages
223 -- 223

Session 5a: Analysis

TBASCEM - Tight Bounds with Arrival and Service Curve Estimation by Measurements:

Authors
Christoph Funda
Thomas Herpel
Reinhard German
Kai-Steffen Jens Hielscher
DOI
10.1145/3629526.3645031
Pages
224 -- 235

Session 5b: Edge

A Learning-Based Caching Mechanism for Edge Content Delivery:

Authors
Hoda Torabi
Hamzeh Khazaei
Marin Litoiu
DOI
10.1145/3629526.3645037
Pages
236 -- 246

Function Offloading and Data Migration for Stateful Serverless Edge Computing:

Authors
Matteo Nardelli
Gabriele Russo Russo
DOI
10.1145/3629526.3649293
Pages
247 -- 257

No Clash on Cache: Observations from a Multi-tenant Ecommerce Platform:

Authors
Anna Lira
Ruan Alves
Thiago Emmanuel Pereira
Fabio Morais
João Ramalho
Mariana Mendes
DOI
10.1145/3629526.3645039
Pages
258 -- 266

MemSaver: Enabling an All-in-memory Switch Experience for Many Apps in a Smartphone:

Authors
Prajwal Challa
Baohua Song
Song Jiang
DOI
10.1145/3629526.3645050
Pages
267 -- 275

Session 6: Micro-services 2

Systemizing and Mitigating Topological Inconsistencies in Alibaba's Microservice Call-graph Datasets:

Authors
Darby Huye
Lan Liu
Raja R. Sambasivan
DOI
10.1145/3629526.3645043
Pages
276 -- 285

Disambiguating Performance Anomalies from Workload Changes in Cloud-Native Applications:

Authors
Alexandru Baluta
Yar Rouf
Joydeep Mukherjee
Zhen Ming Jiang
Marin Litoiu
DOI
10.1145/3629526.3645046
Pages
286 -- 297