KB5122882 and Intel SATA RAID boot problem

A disk read error occurred on a second restart, with the array healthy from an ISO, usually means the boot path (the boot sector, the boot manager, or the storage driver loaded at boot) is what the update changed, not the data. All three boards are older Supermicro boards using Intel SATA RAID, which points at the controller side. What I would check, from the recovery environment (boot your Server ISO, Repair your computer, Command Prompt):

• Confirm the Windows volume is intact and find the right drive letters (dir on each letter, and diskpart > list vol). Then run chkdsk on the system volume. If chkdsk is clean and the array is fine, the files are not the problem.
• Uninstall the update offline, which is the fastest way to get the server back. Run dism /image:D:\ /get-packages and note the package for KB5122882, then dism /image:D:\ /remove-package /packagename:PACKAGE (use the real drive letter and name). If the update was left half applied, run dism /image:D:\ /cleanup-image /revertpendingactions first and restart. Do this on one server first.
• Check the storage driver the boot uses. With Intel RAID the boot depends on the Intel RST or RSTe driver (iaStorA or iaStorV), and an update that replaces or reorders storage drivers can leave the controller without the right one. From the offline registry, check the Start value for the service under SYSTEM\CurrentControlSet\Services (it must be 0 for the boot controller), and load the Intel RSTe driver for your chipset during recovery if the array is not seen under the normal boot.
• Check the BIOS or option ROM. Make sure the RAID option ROM and BIOS are on the latest version for each board, and note whether the server boots in legacy or UEFI mode, because the message you quote is a legacy-boot style message and boot manager changes behave differently in each mode.

Until you know the cause, stop the update on the other machines with the same board and RAID setup. In WSUS or Windows Update for Business decline or defer KB5122882 for that hardware group, patch a spare or a VM first, and report your findings to Microsoft support so the three-board reproduction reaches the right team.