Jetson Nano and PyOpenCL question

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Hello,

i have just installed PyOpenCL on the Jetson Nano.

Quite simple : sudo apt-get install python3-pyopencl

I made a very first test (basic arithmetic operation on 2 huge numpy arrays).

The result :
Jetson Nano + pyopencl : 0.20 seconds
Jetson Nano + pure Numpy : 0.75 seconds

Acceleration factor is about 3.7x

It’s interesting because i previously made tests with pycuda and in that case, pure numpy basic operations are faster than pycuda routines (most of time).

Everything is ok but there is something strange :

When i want to know the platforms i can use with Jetson Nano and pyopencl (pyopencl.get_platforms()), i just get one platform :
[<pyopencl.Platform ‘Portable Computing Language’ at 0x7f6e917008>]

Therefore, i can only get this device :
[<pyopencl.Device ‘pthread-cortex-a57’ on ‘Portable Computing Language’ at 0x19ae7600>]

I made tests with my laptop (i7-8850H CPU and GTX1060 GPU) and i was able to get 2 platforms (GPU and CPU) and 3 devices (GTX1060, INTEL UHD630 and CPU).

When i used my test program with Jetson Nano, i thought i will get 2 platforms (Maxwell GPU and arm cortex a57 CPU).

So, my question is : is it normal to only get Arm cortex a57 device with pyopencl.

Many thanks for your highlights.

Alain

2

Hi Alain, OpenCL isn’t supported on Tegra GPU, hence you only see the CPU listed.

3

Hello dusty,

Many thanks for the answer.

So everything is normal. I will use opencl to boost the basic filters i use with my software and keep pycuda for complex filters.

Have a nice day.

Alain

4

Hello easybob,
did you install only PyOpenCL on Jetson Nano or you had to install also some distribution of OpenCL (for example poCL) ? If yes please tell me which one it was. Thanks in advance.

5

Hello,

i just installed pyopencl, nothing else.

Alain

6

ok thanks, it works