Overview of Performance Hotspots - Overview of Performance Hotspots - 57368

uProf User Guide
Document_ID
57368
Release_Date
2025-06-09
Revision
5.1 English

When the translation is complete, the SUMMARY page will be populated with the profile data and Hot Spots screen will be displayed. The SUMMARY page provides an overview of the hotspots for the profile session through various screens such as Hot Spots and Session Information.

In the Hot Spots screen, hotspots will be displayed for functions, modules, process, and threads. Processes and threads will be displayed only if there are more than one.

The following figure shows the Hot Spots screen:

Figure 1. Summary - Hot Spots Screen

In the above Hot Spots screen:

  • The top 5 hottest functions, processes, modules and threads for the selected event are displayed.
  • The Hot Functions pie chart is interactive in nature. You can click on any section and the corresponding function's source will open in a separate tab in the SOURCES page.
  • The hotspots are shown per event and the monitored event can be selected from drop-down in the top-right corner. You can change it to any other event to update the corresponding hotspot data.
  • From the Select Summary View drop-down, select one of the following:
    • Hot Threads
    • Hot Processes
    • Hot Functions
    • Hot Modules

Based on the selection, one donut will be displayed at a time.

Summary Overview

Based on the selection, the Summary Overview screen will look similar to the following:

Table 1. Summary Overview
Data Collected Table Present Description Timing Details
OS Trace Schedule Summary Summary of per thread running/wait time (percentages).
  • Profile Duration
  • Parallel Time
  • Serial Time
  • Wait Time
  • Sleep Time
Wait Object Summary Time spent in operations related to several types of synchronization objects, that is, locks, mutexes, condition variables, and so on.
Wait Function Summary Time spent in several types of pthread blocking functions, that is, pthread_join, and so on.
Syscall Summary Time spent in syscall(s)
GPU Trace GPU Kernel Summary Time spent per GPU kernel in execution in the enqueued device.
  • Profile Duration
Data Transfer Summary Time spent in GPU data copy operations.
MPI Trace

MPIP2P API

Summary		 Time spent in various MPI P2P API across all ranks of the profile.
  • Profile Duration
  • Parallel Time
  • Serial Time
  • MPI Time
MPI Collective API Summary Time spent in various MPI collective communication API across all ranks of the profile.
CPU Profile Hot Functions Hottest functions based on CPU profile.
  • Profile Duration
  • Parallel Time
  • Serial Time
Hot Modules Hottest modules based on CPU profile.
Hot Threads Hottest threads based on CPU profile.
Hot Processes Hottest processes based on CPU profile.

OS Trace Screen

Figure 2. OS Trace Screen

GPU Trace Screen

Figure 3. GPU Trace Screen

MPI Trace Screen

Figure 4. MPI Trace Screen

CPU Profile

The CPU Profile screen is similar to the Summary - Hotspots Screen.