Using Arm Neon for AI Acceleration on S32G399A-RDB3

[MohamedAbdelslam]

Hi all,

I am working with the S32G399A-RDB3 board and would like to use the Arm Neon SIMD feature to accelerate AI model , as mentioned in the attached screenshot (e.g., 83.2 GFLOPS for lightweight AI tasks).

Could you please provide guidance or documentation on how to enable and utilize the Arm Neon feature for AI acceleration on this platform?

[chenyin_h]

Hello, @MohamedAbdelslam 

Thanks for the reply.

From community supporting perspective, I am not able to find correct one to contact with. For your status, I have the following suggestions:

1. May I know where you purchase the S32G SOC/board? if from your local distributor, I suggest firstly contacting the distributor, for them to provide the information of accessing premium packages.

2. If no information from your distributer, you may also try submitting a private ticket via support.nxp.com, to apply for accessing premium packages.

BR

Chenyin

[chenyin_h]

Hello, @MohamedAbdelslam 

Sorry for the error in link, you may access it via the following link:

https://www.nxp.com/design/design-center/software/eiq-auto-ml-sw-environment:EIQ-AUTO   

For your questions, Yes, there are official AI packages released including examples and document, however, they are premium software and cannot be downloaded directly.

If you are interested in it, I suggest contacting your NXP representative/FAE for details.

BR

Chenyin

[MohamedAbdelslam]

Hello @chenyin_h,

Thank you for the clarification and for providing the correct link.

Could you please let me know how I can get in touch with an NXP representative or FAE regarding the EIQ-AUTO packages? Any specific contact details or channels you recommend would be greatly appreciated.

[chenyin_h]

Hello, @MohamedAbdelslam 

Thanks for the post.

For AI support on S32G platform:

1. Since the Yocto RFS and Ubuntu are included in the BSP release, the customer could try integrating different AI framework like Tensorflow lite to the RFS, and use it with your own applications.

2. From NXP's perspective, you may reference the formal AI support on S32G from 

https://www.nxp.com/design/design-center/software/eiq-auto-ml-sw-environment:EIQ-AUTO 

However, it is premium software package, which could not be directly downloaded, I suggest contacting your NXP representative/FAE, to obtain the way for accessing such packages.

I apologize for your inconvenience.

BR

Chenyin

[MohamedAbdelslam]

Could you please also provide the correct link or access instructions for the eIQ Auto ML package mentioned earlier?
the link currently results in a “Not found” error.

Is there any official example, guide, or sample code from NXP showing how to:

• Enable or leverage NEON on S32G3,

• Optimize AI workloads using NEON