i.MX93 and Linux BSP/eIQ for Automotive

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

i.MX93 and Linux BSP/eIQ for Automotive

Jump to solution
862 Views
ChristophF
Contributor II

I was given the task to clarify some basic questions regarding i.MX93 for the automotive market, that i was not able to clarify by the documentation, faq and the forums alone.

1. The i.MX93 is also marketed as an automotive variant, so is the NXP Automotive Linux BSP which is Automotive-SPICE (tailored for open source), IATF16949 and ISO9001 compliant, also available for the i.MX93?

2. I also saw that there is an ASPICE compliant automotive version of eIQ from NXP, is it possible to use this on i.MX93?

3. Is there any certification for i.MX93 for functional safety or is this processor just not the right pick for functional safety and we should wait for i.MX95 instead?

4. Is it possible to run the NXP eIQ inference engines for the i.MX93 on Linux only or could we choose to use a QNX or Autosar Adaptive instead and still run the eIQ inference?

Thank you for your help,

Christoph

 

Labels (1)
0 Kudos
Reply
1 Solution
845 Views
Zhiming_Liu
NXP TechSupport
NXP TechSupport

Hi @ChristophF 

Q1: It doesn't support Automotive-SPICE IATF16949 or ISO9001 compliant.

Q2: No, the eIQ Auto only support S32 Automotive MCU/MPUs.

Q3: No functional safety for i.MX93. Customer should wait for i.MX95 for functional safety.

Q4: QNX or Autosar Adaptive will not support it. If customer can integrate the ethosu driver from Linux BSP, it's still possible to run ML models on NPU in systems.

Best Regards

Zhiming

View solution in original post

2 Replies
826 Views
ChristophF
Contributor II

Hi Zhiming,

thank you for your reply. This was extremely helpful.

0 Kudos
Reply
846 Views
Zhiming_Liu
NXP TechSupport
NXP TechSupport

Hi @ChristophF 

Q1: It doesn't support Automotive-SPICE IATF16949 or ISO9001 compliant.

Q2: No, the eIQ Auto only support S32 Automotive MCU/MPUs.

Q3: No functional safety for i.MX93. Customer should wait for i.MX95 for functional safety.

Q4: QNX or Autosar Adaptive will not support it. If customer can integrate the ethosu driver from Linux BSP, it's still possible to run ML models on NPU in systems.

Best Regards

Zhiming