afaict Realtek IC isn’t directly connected to this rail (sorry i forgot)
Did you notice any other skewed readings on your other rails earlier?
If not then this is going to be a hard one to track down as the current draw on this high short will be too low to detect it with heat.
There are so many things connected to this rail, for example mosfets on the rear of the board, other main IC’s, the main PMIC etc etc. If the culprit is a capacitor and provided this rail is definitely isolated from all other critical rails i mentioned earlier, then you can connect a bench PSU up to 3V3PDR externally, set your PSU up to supply 3V with a current limit of 500mA (in case soemthing changes) and submerge the board into a tub of IPA, if it’s a failed cap (99% are cracked in some way) then you will see the bubles coming off it.
Failing that, you’ll have to use a milliohm meter and go round the board and try to find the area on this rail which constitues the lowest reading, once you’ve got it you can start pulling components in this immediate area to confirm.
Another way is to use the mV mode of your meter… provided it has the resolution.
And another way is to set your hot air to 100C, connect some fly leads upto the offending short with your meter connected and then hover over certain areas and see which area constitutes the biggest change in ressistance to ground when heat is applied for a few seconds IE: relieving or closing the short as a result of the heat… though i will say this method is a lot less accurate and heat will affect the readings regardless of what area you hovering your hot air, you have to look at the rate of change.
some sort of short sniffer device might be ideal in this scenario… though i’ve never used one and can’t vouch for how effective they are… they might be as useless as a thermal cam for all i know.