• State
  • Locked Locked
  • Replies 0 replies
  • Subscribers 99 subscribers
  • Views 150 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VH-Q1: MSMC Configuration for Each Layer

Kevin Peng
Expert 4628 points

Part Number: TDA4VH-Q1
Other Parts Discussed in Thread: TDA4VH

Tool/software:

Hi TI Experts,

Customer is working on TDA4VH SDK8.5. 

Customer knows how to configure the MSMC for the entire model as we suggested in the below link.

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1377497/tda4vm-understanding-the-use-of-l3-msmc-ram-in-tidl/5310094

However, customer wants to know how to configure the MSMC for each layer.

The reason to ask this question is that, customer currently keep the combined input and output size of the custom operator within 2.9MB. However, the output tensor of the custom operator still cannot fit into the MSMC. But if customer significantly reduce the sizes—for example, setting the input to 0.3MB and the output to 1.3MB—the output tensor can then be placed into the MSMC.

Hence, could you let customer know the below two points please?

  • What is the restriction of using custom operators with MSMC?
  • Any feasibility to configure MSMC usage for each layer?

Customer urgently needs your feedback as soon as possible.

Thanks,

Kevin