Really it wasn't; a TPM is not supposed to cause such issues (i.e. there is no specific combination of options that "must be avoided").

Should I have known what TPM 2.0 was?

Before changing any option you should have looked into what the option does, at least in broad terms.

For example, does it only enable something new, or does it change something from one mode to another? Does it enable a whole new device, or just a mode in an existing device? What is the purpose of that option? Is it widely known that it would conflict with something else?

• The TPM is a separate internal device that is normally passive but can be used by the firmware and/or by the OS (which has drivers for the TPM like for other devices).

• While some systems have a discrete TPM (a completely separate chip) that the option would simply enable/disable, many other systems emulate the TPM through a special mode in the CPU with help from the system's firmware – also known as an "fTPM". Intel PTT is an example of an fTPM; AMD has its own.

• Features that are emulated through system firmware have a greater chance of causing trouble, such as the Ryzen fTPM causing frequent CPU stalls (they released a firmware upgrade to fix that).

• The purpose of a TPM is, mainly, to store keys in a way that they're permanently bound to the system and cannot be exported from it. Therefore it makes very little sense for a "To Go" portable OS to be using the host machine's TPM for anything.

Really it wasn't; a TPM is not supposed to cause such issues (i.e. there is no specific combination of options that "must be avoided").

Should I have known what TPM 2.0 was?

Before changing any option you should have looked into what the option does, at least in broad terms.

For example, does it only enable something new, or does it change something from one mode to another (like how enabling AHCI would disable IDE-emulation)? Does it enable a whole new device, or just a mode in an existing device? What is the purpose of that option (i.e. why does Windows need it)?

• The TPM is a separate internal device that is normally passive but can be used by the firmware and/or by the OS (which has drivers for the TPM like for other devices). It doesn't normally change the basic operation of the system.

• While some systems have a discrete TPM (a completely separate chip) that the option would simply enable/disable, many other systems emulate the TPM through a special mode in the CPU with help from the system's firmware – also known as an "fTPM". Intel PTT is an example of an fTPM; AMD has its own.

• Features that are emulated through system firmware have a greater chance of causing trouble (see also: "SMM mode"), one specific example being the AMD Ryzen fTPM implementation causing frequent CPU stalls (they released a firmware upgrade to fix that). Having "watchdog timeout" indications from the OS might be a sign of fTPM bugs.

• The purpose of a TPM is, mainly, to store keys in a way that they're permanently bound to the system and cannot be exported from it. Therefore it makes very little sense for a "To Go" portable OS to be using the host machine's TPM for anything, although it may be for the purpose of verifying Secure Boot, but practically nothing else as it contradicts the purpose of a portable OS. Most likely the requirement would not be present in official Windows 11 "To Go" images, if such a thing existed.