Python parallel debugging comes to HPC thanks to TotalView upgrade
TotalView has supported C/C++ and Fortran. With the latest quarterly release, it supports Python as well
The OG debugger for supercomputer systems, TotalView, has teamed up with Python, a popular data-science programming language increasingly used in high-performance computing (HPC). At the SC26 supercomputing show, Perforce Software showcased the updated TotalView parallel debugging tool, which now supports Python alongside its existing support for C/C++ and Fortran.
TotalView specializes in dissecting parallel applications for developers, originally designed in the late 1980s for military applications and later acquired by Perforce in 2019. With its latest update, TotalView can handle Python's parallel programming, making it accessible for large, distributed applications in HPC. While Python has gained popularity in HPC due to its ease of learning and use, debugging can become challenging when multiple interpreters run in parallel.
TotalView's tool uses Microsoft's debugpy engine to provide a hierarchical view of hundreds of stack traces, helping developers pinpoint issues. Currently, TotalView supports mpi4py and Python's multiprocessing module, with plans to include other Python concurrency frameworks like Dask and PyTorch's torchrun. The debugger is priced on a token-based model, with each token allowing developers to debug a process.
Written by urgent.news from The Register Science's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.