cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Adventurer
Adventurer
269 Views
Registered: ‎06-14-2018

Multiple ML models on multiple DPU cores

Hi everyone,

Is Xilinx DPU able to run multiple ML models on multiple DPU cores (2 separated models on 2 DPU cores, one model on 1 DPU core) concurrently?

Thank you.

0 Kudos
4 Replies
Highlighted
Xilinx Employee
Xilinx Employee
210 Views
Registered: ‎11-05-2019

Re: Multiple ML models on multiple DPU cores

 

Hello @nvl1109 

 

You can see the Multi-Task demo : Segmentation + Detection.

UG1354 - Vitis AI Library User Guide (v1.0)

 

Thank you

Yoichi

========================================================================
Please mark the Answer as "Accept as solution" if the information provided is helpful.
Give Kudos to a post which you think is helpful and reply oriented.
========================================================================

 

 

Highlighted
Adventurer
Adventurer
178 Views
Registered: ‎06-14-2018

Re: Multiple ML models on multiple DPU cores

Thank you @katsuki for the great info. This document looks like first release from 2019.2. My project currently using 2019.1 Xilinx tools. So is it possible to run similar demo on 2019.1? Thank you.

0 Kudos
Highlighted
Xilinx Employee
Xilinx Employee
163 Views
Registered: ‎11-05-2019

Re: Multiple ML models on multiple DPU cores

 

@nvl1109 

 

Unfortunately that 2019.2 is required to use Vitis AI.

If you want to try it, we recommend using the sample.

 

Thank you

Yoichi

========================================================================
Please mark the Answer as "Accept as solution" if the information provided is helpful.
Give Kudos to a post which you think is helpful and reply oriented.
========================================================================

Highlighted
Adventurer
Adventurer
149 Views
Registered: ‎06-14-2018

Re: Multiple ML models on multiple DPU cores

Thank you. So, the answer could be not possible on 2019.1 Xilinx tools. But on Vitis AI (2019.2), it is possible. ??

0 Kudos