# Xavier multi thread in gpu **URL:** \ **Category:** Jetson AGX Xavier\ **Tags:** cuda\ **Created:** [June 17, 2020, 3:39am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198 "2020-06-17T03:39:41Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![roikyoun](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@roikyoun](https://forums.developer.nvidia.com/u/roikyoun)\ **Post date:** [June 17, 2020, 3:39am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/1 "2020-06-17T03:39:41Z")
Hi NVIDIA is there have any way to use kernel in parallel ? I saw the xavier architecture had several SM. and i want to know the way to run different kernel function attach different SM for run the two kernel at the same time. thanks. ---
**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\ **Post date:** [June 17, 2020, 7:12am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/6 "2020-06-17T07:12:31Z")
Hi, YES. Please check this sample for the concurrent kernel execution: > /usr/local/cuda-10.2/samples/6\_Advanced/concurrentKernels Thanks. ---
**Author:** ![roikyoun](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@roikyoun](https://forums.developer.nvidia.com/u/roikyoun)\ **Post date:** [June 19, 2020, 6:09am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/7 "2020-06-19T06:09:19Z")
Thank you for fast reply currently i’m using Jetson Xavier and trying to use visual profiler via nsight system. run the code in Xavier first and hold some second. and attach the code by PID for checking GPU function runs well. but i can not understand the result. name of the kernel function is all as number. like GPU context : 102XXX … CPU PID : 20045 … is there have way to know about GPU context name as words ? thanks !! ---
**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\ **Post date:** [June 19, 2020, 6:55am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/8 "2020-06-19T06:55:31Z")
Hi, Suppose the profiling data also has the executed function name. You can find some information from it. An alternative is to print out the process ID from your program so you can match it to the Nsight. Thanks. ---
**Author:** ![roikyoun](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@roikyoun](https://forums.developer.nvidia.com/u/roikyoun)\ **Post date:** [June 19, 2020, 7:53am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/9 "2020-06-19T07:53:33Z")
thank you for fast reply !! sorry i can’t understand about "print out the process ID form your program to match to the Nsight " fallowing pic is the result of nsight system. for running /usr/local/cuda-10.2/samples/6\_Advanced/concurrentKernels this codes . as you see in this picture name of GPU function are all in number. ![result](https://global.discourse-cdn.com/nvidia/original/3X/3/0/3075bfde5b4faec6c436697312538796619d2ea1.png) could you explain more detail about print out process ID from program(?) to know name of kernel function. thanks !! ---
**Author:** ![roikyoun](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@roikyoun](https://forums.developer.nvidia.com/u/roikyoun)\ **Post date:** [June 23, 2020, 1:58am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/10 "2020-06-23T01:58:42Z")
Is there anyone here ? ---
**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\ **Post date:** [July 8, 2020, 6:33am UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/11 "2020-07-08T06:33:52Z")
Hi, Sorry for the late update. It looks like you are trying _Nsight System_ rather _Nsight_. Could you give Nsight tool a try? > **[Nsight Eclipse Edition](https://developer.nvidia.com/nsight-eclipse-edition)** > > The standalone version of Nsight Eclipse Edition has been discontinued As of CUDA Toolkit 11.0, the standalone version of Nsight Eclipse Edition has been discontinued in favor of a new set of plugins: Nsight Eclipse Plugins for the Eclipse IDE.... You should be able to get the profiling data as the section in **Nsight Profiler** Thanks. ---
**Author:** ![TomNVIDIA](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/tomnvidia/32/14181_2.png) [@TomNVIDIA](https://forums.developer.nvidia.com/u/TomNVIDIA)\ **Post date:** [October 18, 2021, 6:22pm UTC](https://forums.developer.nvidia.com/t/xavier-multi-thread-in-gpu/128198/13 "2021-10-18T18:22:12Z")