Intel® High Level Design
Support for Intel® High Level Synthesis Compiler, DSP Builder, OneAPI for Intel® FPGAs, Intel® FPGA SDK for OpenCL™
Announcements
The Intel sign-in experience is changing in February to support enhanced security controls. If you sign in, click here for more information.

openCL memory map

paolop
Beginner
548 Views

Hi, I am trying to interface a design (gzip) implemented through the OpenCL flow. I went successfully through all the steps, till generating the green bitstream which ran properly on a board equipped with Arria 10. Now I need to interface the design without using the OpenCL layer, so I need to discover where the kernels and the memory sections are mapped: I need to be able to
- program DMAs to transfer data to/from the card,
- give the start to the various kernels,
- exchange scalar parameters with the kernels and
- be informed when some events happen in the FPGA (end of a transfer, stop of a kernel).
As I did not find such information, could you please give me a reference where I could find them?
FYI, I should put the gzip kernel together with another AFU which has been already developed in the OPAE env; also gzip has been developed in the same OPAE env, but we are not able to access it because we cannot find the address map. When I say " developed in the OPAE env", I mean that I took the files generated by openCL for the gzip kernel and I moved them within the project of the other AFU, developed in OPAE env.
I hope that I was able to clarify what I need.

0 Kudos
5 Replies
aikeu
Employee
473 Views

Hi paolop,


I will get back to you when I have revelant info to your question.


Thanks.

Regards,

Aik Eu


paolop
Beginner
455 Views

Thanks for taking care of this question. It is related to a PoC made in collaboration with Intel PSG Group, and we must achieve the best performance possible as soon as possible.
The question is relevant to reaching the goal.
Thanks

Emilio Billi

 

 

aikeu
Employee
429 Views

Hi paolop,


I get the info from the team that what you are trying to work on is to convert the code from OpenCL to RTL coding which is not supported.

It is required to use the OpenCL kernel instead or else it is required to rewrite the AFU using RTL only can control the DMA and etc.


Do let me know if you have any question from the above statement.


Thanks.

Regards,

Aik Eu


aikeu
Employee
396 Views

Hi paolop,


I will close this thread if no further quesiton.


Thanks.

Regards,

Aik Eu


aikeu
Employee
281 Views

Hi paolop,


We do not receive any response from you to the previous question/reply/answer that I have provided. This thread will be transitioned to community support. If you have a new question, feel free to open a new thread to get the support from Intel experts. Otherwise, the community users will continue to help you on this thread. Thank you.


Thanks.

Regards,

Aik Eu


Reply