SLIDE 1
Technische Universität München
Performance analysis with Periscope
- M. Gerndt, V. Petkov,
- Y. Oleynik, S. Benedict
Technische Universität München
Performance analysis with Periscope M. Gerndt, V. Petkov, Y. - - PowerPoint PPT Presentation
Performance analysis with Periscope M. Gerndt, V. Petkov, Y. - - PowerPoint PPT Presentation
Technische Universitt Mnchen Performance analysis with Periscope M. Gerndt, V. Petkov, Y. Oleynik, S. Benedict Technische Universitt Mnchen September 2010 Technische Universitt Mnchen Outline Motivation Periscope architecture
SLIDE 2
SLIDE 3
Technische Universität München
Common performance analysis procedure on Power6 systems
Use Tprof to pinpoint time-consuming subroutines Use Xprofiler to understand call graph; mpitrace for MPI comm Use hpmcount (libhpm) to measure HW Counters
Problem
Routine, error-prone and time-consuming Requires deep HW knowledge Mostly post-development process Hard to map bottlenecks to their source code location
Solution
Automate the performance analysis Integrate parallel application development and performance analysis within the same IDE
Motivation
SLIDE 4
Technische Universität München
Periscope
Iterative online analysis
Measurements are configured, obtained and evaluated on the fly no need to store trace files
Distributed architecture
Reduced network overhead Analysis performed by multiple distributed hierarchical agents
Automatic bottlenecks search
Based on performance optimization experts' knowledge Single-node Performance on Intel Itanium6, IBM Power6, x86-s MPI Communication OpenMP Performance
Enhanced Eclipse-based GUI Instrumentation: Fortran, C/C++; MPI / OpenMP / Hybrid
SLIDE 5
Technische Universität München
Distributed architecture
Graphical User Interface Application Interactive frontend
Eclipse-based GUI Analysis control Agents network Monitoring Request Interface
SLIDE 6
Technische Universität München
Candidate Properties
Proven Properties
Analysis
Performance Measurements
Refinement
Raw performance data
Instrumented Application Analysis Agents GUI
Precision Location
Monitoring Requests
Start
Final Properties Report
SLIDE 7
Technische Universität München
Automatic search for bottlenecks
Automation based on formalized expert knowledge
Efficient search algorithms strategies
Performance property
Condition Confidence Severity
Performance analysis strategies
Itanium2 Stall Cycle Analysis IBM POWER6 Single Core Performance Analysis MPI Communication Pattern Analysis Generic Memory Strategy OpenMP-based Performance Analysis Scalability Analysis OpenMP codes
SLIDE 8
Technische Universität München
POWER6 Single Core Performance Properties
Hot spot of the application
Cycles lost due to cache misses Average amount of cycles lost per L1 miss High L1 demand load miss rate High L2 demand load miss rate High L3 demand load miss rate Cycles lost due to address translation misses Cycles lost due to store instructions Cycles lost due to Floating Point instructions inefficiencies Cycles lost due to Integer multiplications and divisions Cycles lost due to no instruction to dispatch
SLIDE 9
Technische Universität München
Itanium2 Stall Cycle Properties
IA64 Pipeline Stall Cycles
Stalls due to pipeline flush Stalls due to branch misprediction flush Stalls due to exception flush Stalls due to floating point exceptions or L1D TLB misses Stalls due to Flush to zero or SIR stalls Stalls due to L1D TLB misses ... Stalls due to waiting for data delivery to register Stalls due to waiting for integer register Stalls due to waiting for integer results Stalls due to waiting for FP register Stalls due to waiting for integer loads L3 misses dominate data access L2 misses L3 misses Stalls due to register stack engine
SLIDE 10
Technische Universität München
MPI Communication Patterns Analysis
MPI_Recv MPI_Send
p1 p2
Automatic detection of wait patterns Measurement on the fly No tracing required!
SLIDE 11
Technische Universität München
MPI Performance Properties
Excessive MPI time in receive due to late sender Excessive MPI time due to late root in broadcast Excessive MPI time in root due to late process in reduce Excessive MPI time in ... (1xN, Nx1, 1x1, NxN) Excessive MPI time due to many small messages Excessive MPI communication time
SLIDE 12
Technische Universität München
OpenMP-based Performance Properties
Searches OpenMP-based perf. problems in a single step Properties are divided into four major domains
Startup and Shutdown Overhead Load Imbalance in OpenMP regions: Parallel region Parallel loop Explicit barrier Parallel sections Not enough sections Uneven sections
- Seq. Computation in parallel regions:
Master region Single region Ordered loop OpenMP Synchronization properties Critical section overhead property Frequent atomic property
OpenMP Tasking analysis under development
SLIDE 13
Technische Universität München
Scalability Analysis OpenMP codes
Identifies the OpenMP code regions that do not scale well Scalability Analysis is done by the frontend No need to manually configure the runs and find the speedup!
Frontend initialization
Frontend.run()
- i. Starts application
ii.Starts analysis agents iii.Receives found properties
n
Extracts information from the found properties Does Scalability Analysis Exports the Properties GUI-based Analysis After 2n runs
SLIDE 14
Technische Universität München
Scalability Analysis Properties
Meta-Properties
Properties occurring in all configurations Property with increasing severity across the configurations
Speedup-based Prop.
Linear Speedup Super linear Speedup Linear Speedup failed for the first time Speedup Decreasing
- Exp. specific properties
Code region with the lowest speedup Low Speedup based on a threshold
SLIDE 15
Technische Universität München
Graphical User Interface
Integrates with the Eclipse Development Platform
Open-source, extensible and very popular IDE Supports different programming languages: C/C++, Fortran, etc. Uses the Eclipse Parallel Tools Platform (PTP) which provides a higher-level abstraction of the underlying parallel system
Designed to combine:
Performance measurement functionality of Periscope Advanced IDE functions like code indexing, refactoring, etc.
Features
Multi-functional table to display the detected bottlenecks Outline of the instrumented code regions Clustering techniques to get classes of similarly behaving processes Supports both local and remote projects Higher-level configuration and execution of performance experiments
SLIDE 16
Technische Universität München
Project view Source code view Properties view SIR outline view
Graphical User Interface
SLIDE 17
Technische Universität München
Simple and clean tree-based overview Multi-level grouping Complex data filtering Multiple criteria sorting algorithm Navigation from the properties to their source code location
Periscope GUI: Properties Table
SLIDE 18
Technische Universität München
Resembles the code outline view of the Eclipse C/C++ Development Tooling Outlines the instrumented code regions and their nesting Shows the number of properties in each region Assists code navigation Filters the displayed properties
Periscope GUI: Instrumentation Outline
SLIDE 19
Technische Universität München
Eclipse File System (EFS)
Abstracts the underlying file system details Any supported file system can be used: Remote projects using SSH/FTP/DStore, Local, Zip, etc. Source files of the analyzed application reside only on the remote no need for synchronization
Remote Development Tools (RDT)
Part of Eclipse Parallel Tools Platform (PTP) Project Remote Compilation Remote Indexing Currently supports only C/C++ applications
Periscope GUI: RDT and EFS
SLIDE 20
Technische Universität München
External Tools Framework (ETFw)
Part of Eclipse Parallel Tools Platform (PTP) Project More convenient environment using ETFw's Profile launch configuration
no terminal access needed higher level configuration and automation possible
Periscope GUI: Experiment Configuration
SLIDE 21
Technische Universität München
Clustering support
Properties summarization
Metaproperties
Identify hidden behavior Weka workbench in the GUI:
Waikato Environment for Knowledge Analysis Uses K-Means algorithm Groups properties based on CPU distribution and code region
Results shown in a table view similar to the properties view Done also on the fly in the hierarchy
Property 1 Property 2 Property 3 Cluster 1 Cluster 2 CPUs: 2-3,5,11,13-14 Cluster 3 Cluster 1 CPUs: 7-10,16 Cluster 3 CPUs:1,4,6,12,15 Cluster 2
SLIDE 22
Technische Universität München