Friday, 2021-08-13













Main Conference Program


Day 3
09:00 --> 18:00

(Mon, 24)

Day 4
09:00 --> 18:00

(Tue, 25)

Day 5
09:00 --> 16:25

(Wed, 26)
09:00 --> 09:15
09:00 --> 10:00

Francesco Quaglia
09:00 --> 10:00

Arif Merchant
09:15 --> 10:15

Thomas Wuerthinger
Workshop reports
10:00 --> 10:15
Workshop reports
10:00 --> 10:15

Session 1
10:45 --> 12:50

Memory and Hardware Impact

Session 4
10:45 --> 12:50

Industry and Experience

Session 7
10:45 --> 12:15

Design for Performance and Problem diagnosis

Presentation of SPEC Research Group
12:15 --> 12:35

12:35 --> 12:50

lunch break

Session 2
14:20 --> 16:10

Online Performance Management

Session 5
14:20 --> 16:15

Analytic Models

Session 8
14:20 --> 16:20

Empirical Studies

Demo Session
16:10 --> 18:00


16:20 --> 16:30

Session 3
16:35 --> 18:00

Model Building

Session 6
16:40 --> 17:55


SPEC RG Annual Meeting
16:45 --> 21:00

RG Cloud, RG DevOps and RG BigData working groups

Welcome Reception Banquet
  • Full paper --> 25 minutes (incl. questions)

Daily Schedules

Session 1: Memory and Hardware Impact (chair: Murray Woodside)

  • A Comprehensive Java Benchmark Study on Memory and Garbage Collection Behavior of DaCapo, DaCapo Scala, (full paper)
  • An Incremental Methodology for Energy Measurement and Modeling(full paper)
  • Detecting Memory-Boundedness with Hardware Performance Counters(full paper)
  • Transferring Performance Prediction Models Across Different Hardware Platforms(full paper)
  • Using Libception to Understand and Improve HTTP Streaming Video Server Throughput [Best Paper Candidate]

Session 2: Online Performance Management

  • An Empirical Analysis of Amazon EC2 Spot Instance Features Affecting Cost-effective Resource Procurement(full paper) [Best Paper Candidate]
  • [Best Paper Candidate]
  • Dynamic memory partitioning for cloud caches with heterogeneous backends(wip paper)
  • Latency Aware Elastic Switching-based Stream Processing Over Compressed Data Streams(full paper)
  • On Moving Averages, Histograms and Time-Dependent Rates for Online Measurement(full paper)

Session 3: Model Building

  • AutoPerf: Automated Load Testing and Resource Usage Pro?ling of Multi-Tier Internet Applications(full paper)[Best Paper Candidate]
  • Deriving Parameters for Open and Closed QN Models of Operational Systems Through Black Box Optimization(full paper)
  • HyperStar2: Easy Distribution Fitting of Correlated Data(wip paper)
  • IRIS: Iterative and Intelligent Experiment Selection (full paper)

Session 4: Industry and Experience(chair: Meikel Poess)

  • A Performance-Centric Approach to Complex Decisions Support - Ate Penders, Gregor Pavlin, Ana Lucia Varbanescu and Henk Sips (full paper)
  • CloudPerf: A Performance Test Framework for Distributed and Dynamic Multi-Tenant Environments(full paper)
  • Cost-Efficient and Reliable Reporting of Highly Bursty Video Game Crash Data(full paper)
  • Early Detection of Performance Anomalies in Large Scale Software Systems(full paper)
  • HPC Supported Mission-Critical Cloud Architecture -Tae Joon Jun, Myong Hwan Yoo, Daeyoung Kim, Kyu Tae Cho, Seungyoung Lee and Kyuoke Yeun (full paper)

(chair:Andre Bondi)

  • A Tool Supporting the Analytical Evaluation of Service Level Agreements(full paper)
  • (full paper)
  • Practical Implication of Analytical Models for SSD Write Amplification(short paper)
  • Schedulability and Memory Interference Analysis of Multicore Preemptive Real-time Systems(full paper)
  • Timeliness Evaluation of Intermittent Mobile Connectivity over Pub/Sub Systems(full paper)

  • Conducting Repeatable Experiments in Highly Variable Cloud Computing Environments(short paper)
  • HTAPBench: Hybrid Transactional and Analytical Processing Benchmark(full paper)
  • TARUC: A Topology-Aware Resource Usability and Contention Benchmark(full paper)
  • Work in Progress : Cloning IO Intensive Workloads Using Synthetic Benchmark(wip paper)

Session 7: Design for Performance and Problem diagnosis(chair: Catia Trubiani)

  • An Efficient Performance-Driven Approach for HW/SW Co-Design - Daniele Di Pompeo, Emilio Incerto, Vittoriano Muttillo, Luigi Pomante and Giacomo Valente (wip paper)
  • (wip paper)
  • Efficient Sampling-based Lock Contention Profiling for Java(wip paper)
  • Identifying Derived Performance Requirements of System Components from Explicit Customer- and Application- Facing Performance Requirements(wip paper)
  • Many Flies in One Swat: Automated Categorization of Performance Problem Diagnosis Results(wip paper)
  • Performance Analysis of Applications in the Context of Architectural Rooflines (wip paper)
  • Performance and Dependability evaluation of distributed event-based systems: a dynamic code-injection approach(wip paper)
  • Predicting Power Consumption of High-Memory-Bandwidth Workloads (Work in Progress)(wip paper)
  • User-centered Offline Analysis of Memory Monitoring Data(wip paper)

Session 8: Empirical Studies

  • An Empirical Study of Computation-Intensive Loops for Identifying and Classifying Loop Kernels(full paper)
  • An Exploratory Study of the State of Practice of Performance Testing in Java-Based Open Source Projects(full paper)
  • Collaborative Computing for Heterogeneous Integrated Systems(wip paper)
  • Empirical Study of Usage and Performance of Java Collections(full paper)
  • Unit Testing Performance in Java Projects: Are We There Yet ?(full paper)
  • Work-in-Progress Paper: Agile Scalability Requirements(wip paper)

  • and Analysis
  • SQL Query Volume Performance Estimation Tool
  • Accelerating Java Streams With A Data Analytics Hardware Accelerator
  • software approach for timing analysis on resource constrained systems
  • Modelling ToolsGiuliano Casale, Mattia Cazzoli, Shuai Jiang, Vitor Lopes, Giuseppe Serazzi and Lulai Zhu
  • Modeling and Prediction
  • Autopilot: Enabling easy Benchmarking of Workload Energy Efficiency
  • Intensive Workloads
  • CloudPerf: A Performance Test Framework for Distributed and Dynamic Multi-Tenant Environments
  • In-Memory Data Grid Should I Use? Benchmarking Hazelcast and Infinispan
  • Developers Environment
  • Online Measurement of Time Series