An MCU doesn't sit in an engine bay, neither does any kind of...

An MCU doesn't sit in an engine bay, neither does any kind of memory chip. The sensors sit there, which send signals back to the MCU.
So this could be a complete misunderstanding from Peter in regards to autoelectrics and EV's. Also 120-200 degrees - engines must avoid reaching that boiling point of water / coolant or you're going to destroy your engine. That boiling point is ~127 degrees. Most engines should be sitting around 90 degrees with appropriate coolant, but thats also not the point because a memory chip won't be sitting in or on the engine - it will be behind the firewall of the vehicle.

Now that we've cleared that up. It may also not be a misunderstanding - being a brand new PCMO RRAM chip which there is nothing to trial and comapre against and the effects of temperature very well could make the ionization more sensitive and lead to loss of data.

I don't have anything data on PrCaMnO3, instead we'll have to refer to a 2008 paper from Samsung that's using NiO instead - "Stack Friendly All-Oxide 3D RRAM using GaInZnO Peripheral TFT realized over Glass Substrates" - https://picture.iczhiku.com/resource/ieee/wHiFzFQePlfKJBVb.pdf

"Figure is the operation of the RRAM cell at hightemperatures. The LRS shows a slight increase with increasingtemperature similar to metallic behavior while the HRS showsincrease of conductivity with increasing temperature similar tosemiconductors"


So theoretically, this should be completely fine if PCMO shares similar characteristics where the resistance isn't decreasing until its past the 100 degree point - which means its also perfectly fine for most embedded memory applications including automotive. So it doesn't make sense for a 4DS representative to say otherwise, unless they know something.

This is why clarification is important.