Cube, which is used as performance report explorer for Scalasca and Score-P, is a generic tool for displaying a multi-dimensional performance space consisting of the dimensions (i) performance metric, (ii) call path, and (iii) system resource.
The Cube Software Framework facilitates the collection, processing, and analysis of application performance profiling data from Score-P and Scalasca. With this framework, users have access to a comprehensive set of tools for measuring and understanding application performance.
Cube (CUBE Uniform Behavioral Encoding) is suitable for analyzing a wide variety of performance data for parallel programs including MPI and OpenMP applications.
Cube has been designed around a high-level data model of program behavior called the cube performance space. The Cube performance space consists of three dimensions: a metric dimension, a program dimension, and a system dimension.
The metric dimension contains a set of metrics, such as communication time or cache misses. The program dimension contains the program's call-tree, which includes all the call paths onto which metric values can be mapped. The system dimension contains the components executing in parallel, which can be processes or threads depending on the parallel programming model. Each point of the space can be mapped onto a number representing the actual measurement for metric while the control flow of process/thread was executing call path. This mapping is called the severity of the performance space.
Cube provides a general purpose library CubeLib to read and write instances of the previously described data model in the form of a cubex file and set of command-line tools for its processing.
Please find the latest tarballs here:
https://perftools.pages.jsc.fz-juelich.de/cicd/cubelib/
If you have any comments or questions regarding the use and installation of CubeLib, or want to report a bug you discovered, please email scalasca@fz-juelich.de
You can also sign up to the Scalasca News mailing list to receive the latest news about new releases, tutorials, workshops, and other Scalasca-related events.
If you find CubeLib helpful for your research, please mention it in your publications:
This work is supported by BMBF, DFG, Helmholtz POF, EU (FP7, Horizon 2020, ITEA-2), EuroHPC JU, US DOE, Siemens AG, Intel.
The Scalasca Trace Tools support performance optimization of parallel programs with a collection of highly scalable trace-based tools for in-depth analyses of concurrent behavior, in particular with respect to communication and synchronization, and offers guidance in exploring their causes.
Score-P provides insight into massively parallel HPC applications, their communication, synchronization, I/O, and scaling behaviour to pinpoint performance bottlenecks and their causes.