E3800 Series VLI89 Bug (System May Experience Inability to Boot or May Cease Operation)

[IngoZ]

We have received 100 PC104 CPU cards from ADLINK company with ATOM CPU E3815 in summer 2018.

It took about 2 years for all CPU cards to be installed and put into operation in the overall system. We already noticed problems with booting during the integration phase. We suspected the cause to be a lack of support for the CPU by Linux. We implemented our own reset mechanism on register level and sorted out CPU cards with boot problems. The overall system went into operation this way.

Now, within a short time, two CPU cards failed again completely due to boot problems.

Thereupon we handed over all failed CPU cards to the manufacturer, together with the request for clarification of the cause of the error.

Currently our overall failure rate is 8%.

The manufacturer's analysis says the following:

1. all CPU cards show the same bug.

2. signal LPC_CLK, signal level is too low. The CPU is defective.

Our own research then leads us to the CPU error VLI89, which has been known since 2017.

This is a very big problem for us, because it turned out that we received CPU cards with the E3815 ATOM CPU, which all have the VLI89 bug.

What has Intel's position been in similar cases?

[Jaime_Lizarme]

Hello @IngoZ ,

 Thank you for contacting Intel Embedded Community.

I would like to inform you that Bay Trail Embedded SKU’s have reached their End of Interactive Support (EOIS).  If you need further details, please contact your Intel product distributor.

Best regards,

Jaime L.

Intel Customer Support Engineer