TDA2SX: How to speed up and optimize initialization load flow of deep learning?

Jiacai Wang

Part Number: TDA2SX

Hi,

How to speed up initialization load flow of deep learning? System spends 20 seconds on initialization load of deep learning.

Deep learning module always consumes over 200M Bytes DDR3 RAM capacity when power up, if we wish system can release RAM capacity while we don't use object detection.

over 6 years ago

0 Gaviraju Doddabettahalli Bettegowda over 6 years ago

TI__Genius 14305 points

1) How to speed up initialization load flow of deep learning? System spends 20 seconds on initialization load of deep learning.

>>>>> refer this thread:

2) Deep learning module always consumes over 200M Bytes DDR3 RAM capacity when power up, if we wish system can release RAM capacity while we don't use object detection.

>>>> I will check with team & get back

0 Jiacai Wang over 6 years ago

Intellectual 420 points

Hi,

System startup time is within 5 seconds except for deep learning load.

Thanks

0 Gaviraju Doddabettahalli Bettegowda over 6 years ago in reply to Jiacai Wang

TI__Genius 14305 points

Hi,

If you don't use object detection algorithm, then remove from the use-case

Thanks

Gaviraju

0 Jiacai Wang over 6 years ago in reply to Gaviraju Doddabettahalli Bettegowda

Intellectual 420 points

Sometimes will use object detection algorithm.

Processors

Processors forum

TDA2SX: How to speed up and optimize initialization load flow of deep learning?