TotalView for HPC
Dynamic Analysis and Debugging of Complex Code
TotalView for HPC is a debugging software designed for high-performance computing (HPC) environments. TotalView enables faster fault isolation, improved memory optimization, and dynamic visualization for your high-scale HPC apps. TotalView understands high-scale parallel and multicore applications, with unprecedented control over processes and thread execution and visibility into program states and data.
Reproduce Difficult Problems Quickly
With TotalView for HPC, simultaneous debug many processes and threads in a single window to get complete control over program execution: Running, stepping, and halting line-by-line through code within a single thread or arbitrary groups of processes or threads. Work backwards from failure through reverse debugging, isolating the root cause faster by eliminating repeated restarts of the application. Reproduce difficult problems that occur in concurrent programs that use threads, OpenMP, MPI, CUDA, and GPUs, or coprocessors. Debug autonomous machine applications running on the NVIDIA Jetson AGX Xavier ARM/Volta GPU platform.
Easily Start Debugging Sessions with TotalView Reverse Connect
TotalView’s new Reverse Connect feature easily allows you to establish a connection where the TotalView UI is running on a front-end node and debugging a job executing on compute nodes. This workflow greatly simplifies establish interactive TotalView debugging sessions in clusters. For more information, see Chapter 18, Reverse Connections, in the TotalView User Guide.
No More Memory Leaks or Deadlocks
Memory leaks, deadlocks, and race conditions can be things of the past. Whether an experienced developer or new to the development challenges of multicore or parallel applications, TotalView finds errors quickly, validates prototypes, verifies calculations, and certifies code correctness.
Debugging in C/C++ and Python
Easily analyze and debug multi-language applications written in both Python and C/C++ by seeing an integrated cell stack and easily examining variables in both sets of languages.
Seamless Integration and Support
TotalView works with C, C++, and Fortran applications written for Linux (including the Cray and Blue Gene platforms), Linux PowerLE, UNIX, Mac OS X, and Xeon Phi, and supports OpenMP, MPI, OpenPOWER, OpenACC / CUDA and ARM.
Los Alamos National Laboratory