"overclocking" an RT10xx

[giuseppescarpi]

Not that I plan to do so... but I wonder if there is any real difference between RT microcontroller that run up to 528 MHz or up to 600 MHz.

What happens if the 528 MHz model is "overclocked" at higher speed? Is it only a matter of shortening the component's life? Or are there other issues?

[gusarambula]

Hello Giuseppescarpi,

Overclocking a part would mean to operate outside of the tested operating parameters, so robust operation cannot be guaranteed. Besides, the lifetime of the component would be reduced, so I would strongly recommend against overclocking any MCU. If you need to operate at 600Mhz the recommendation would be switching to a 600Mhz part number.

My apologies for the inconvenience.

Regards,
Gustavo

[mastupristi]

Overclocking a part would mean to operate outside of the tested operating parameters, so robust operation cannot be guaranteed. Besides, the lifetime of the component would be reduced, so I would strongly recommend against overclocking any MCU. If you need to operate at 600Mhz the recommendation would be switching to a 600Mhz part number.

I am absolutely agree with you to not exceeding the specified limits under normal conditions. But right now, we're not experiencing a normal time.

Our distributor has delayed all orders placed for the past six months by rescheduling deliveries to August 2022 at the earliest.

This pushes us to be more "elastic" in our beliefs.

In any case your answer only focuses on shortening the life of the component, or is that wrong?

We checked the AN12170 i.MXRT1050 Product Lifetime Usage Estimates.

The industrial component @528MHz @95 degrees has a PoH 160000 = 18.2 years.
The commercial component @528MHz @95 degrees has a PoH 76000 = 8.6 years.
The commercial component @600MHz @95 degrees has a PoH 28000 = 3.2 years.

The industrial component @528MHz @80 degrees has a PoH 400000 = 45.6 years.
The commercial component @600MHz @80 degrees has a PoH 72000 = 8.2 years.

As you can see industrial silicon is better than commercial even under the same conditions.
So if the issue is component life, you need to compare with the life expectancy of the commercial component.

Our application will operate at temperatures far below those indicated. We rarely exceed 60 degrees, far from the maximum operating limits

Our FAEs have repeatedly told us that commercial and industrial components come from the same wafers, so they're the same mask. The selection is made in the first instance based on the position of the component in the wafer. Is this correct?

So besides shortening the life, what would it mean to make an industrial component work @600MHz? Is it really working out of spec assuming the silicon (the mask) is the same (but a better selection) as the commercial component?
Isn't it that the problem is only commercial and not technical?

best regards

Max