CUDA Profiler Tools Interface (CUPTI) for CUDA Toolkit 11.8 is now available

Developer Tools CUPTI – CUDA Profiler Tools Interface
1

CUDA Profiler Tools Interface (CUPTI) for CUDA Toolkit 11.8 is now available for download in the NVIDIA Registered Developer Program.

The NVIDIA® CUDA Profiling Tools Interface (CUPTI) is a dynamic library that enables the creation of profiling and tracing tools that target CUDA applications. CUPTI provides a set of APIs targeted at ISVs creating profilers and other performance optimization tools:

  • the Activity API,
  • the Callback API,
  • the Event API,
  • the Metric API,
  • the Profiler API,
  • the PC Sampling API, and
  • the Checkpoint API

Using these CUPTI APIs, independent software developers can create profiling tools that provide low and deterministic profiling overhead on the target system, while giving insight into the CPU and GPU behavior of CUDA applications.

CUDA Profiler Tools Interface (CUPTI) for CUDA Toolkit 11.8 includes these improvements and updates:

New Features

    CUPTI has made the following changes as part of the CUDA Toolkit 11.8 release:
    • CUPTI adds tracing and profiling support for devices with compute capability 9.0 i.e. NVIDIA H100 GPUs and systems that are based on H100 GPUs.
    • Added new fields clusterX, clusterY, clusterZ, and clusterSchedulingPolicy to output the Thread Block Cluster dimensions and scheduling policy. These fields are valid for devices with compute capability 9.0 and higher. To accomodate this change, activity record CUpti_ActivityKernel7 is deprecated and replaced by a new activity record CUpti_ActivityKernel8.
    • A new activity kind CUPTI_ACTIVITY_KIND_JIT and corresponding activity record CUpti_ActivityJit are introduced to capture the overhead involved in the JIT (just-in-time) compilation and caching of the PTX or NVVM IR code to the binary code. New record also provides the information about the size and path of the compute cache where the binary code is stored.
    • PC Sampling API is supported on Tegra platforms - QNX, Linux (aarch64) and Linux (x86_64) (Drive SDK).

Resolved Issues

    CUPTI has made the following fixes as part of the CUDA Toolkit 11.8 release:
    • Resolved an issue that might cause crash when the size of the device buffer is changed, using the attribute CUPTI_ACTIVITY_ATTR_DEVICE_BUFFER_SIZE, after creation of the CUDA context.
## Requirements **For more information on CUPTI for CUDA Toolkit 11.8** , including features, requirements, documentation and support, please visit the **[CUPTI Overview page ](https://developer.nvidia.com/CUPTI-CTKCUDA Profiler Tools Interface (CUPTI) for CUDA Toolkit 11_8)** .

To download this version, get it as part of the CUDA Toolkit CUDA Profiler Tools Interface (CUPTI) for CUDA Toolkit 11.8 .
Version 2022.3.0 Overview & New Features | Download | Documentation

Latest PRODUCT INFO