I made a cringe-worthy mistake a few days ago. The Wraith Spire cooler my CPU came with had more than usual amount of thermal paste on it. So, when I tried to replace the thermal paste, I saw that it had oozed on the sides of the socket. I took a cotton swab, dabbed it in some IPA and cleaned it off. Unfortunately, I didn't see a thin strand of paste that was on the side of the socket. And, as luck would have it, just when i took the CPU out of the socket, it got on one of the pins. So I took out one of my new painting brush with the finest bristles, dabbed it in IPA and cleaned the pin. When I was done, it was clean. I couldn't see any more paste even with a magnifying glass. A tiny blob of paste also fell on the motherboard just above the socket, so I cleaned that as well with the brush. I reseated the CPU, cleaned the cooler, applied a fresh paste of thermal compound, and all was well.
It has been running fine, but I noticed a change while monitoring the voltages. I use an Asus Strix X370-F motherboard with latest BIOS. I had the VDDCR CPU Voltage set to 1.175V. It has always run at that voltage. I have not overclocked my system yet. Anyway, in HWMonitor, I used to see CPU VCore voltage drop to about .875V sometimes. But the maximum was close to 1.175V. Sometimes it crossed 1.175V by just a bit, like 1.188V, but the thing is, when the system was idle, the voltage always dropped down. Now, after the thermal paste covfefe, the CPU VCore refuses to drop down. It's always at 1.177V, even at idle. Also, even though the CPU VDD is at 1.175V, the VID voltages have shot up to 1.375V. Is that normal?
Does this mean the thermal paste has partially damaged the pins of the CPU or the motherboard socket?
A screenshot of HWMonitor is attached.
I am kinda-not-really overclocking (I just run it at the Turbo frequency all the time) and that's what mine runs at
Have you set your VCore manually? And if you have, does HWMonitor show any drops in CPU VCORE when the system is idle?
I use P-state overclocking so only the highest performance state has increased voltage, the lower P states, which it was in when this screenshot was taken, uses the proper 0.9v. I don't particularly LIKE HWMon for voltage monitoring since it can give incorrect readings, obviously my Vcore is not 0.676v under full speed, I prefer AIDA64 (above) or ASUS AI Suite (bottom).
Umm. I think you forgot to attach AIDA64. Anyway, my motherboard Asus Strix X370F had the P-State Overclocking option in BIOS, which was, surprise surprise, removed in later BIOS updates. If I knew that, I'd have gone with Asus Crosshair VI or Asrock Taichi.
Sorry. My bad.
In my case, even Ryzen Master or Asus AI Suite 3 reports a near-constant VCore even when the system is idle. So, HWMonitor is not at fault here.
Another thing to note is that Ryzen processors are able to change voltages -very- rapidly, helped by Windows 10's ability to do that, much faster than Bulldozer derived processors are able to. Also, some programs -still- have problems reading the core voltage, CPUz is a prime example, it is not using a mere 0.2v on idle and 0.665v on full load, so it's also a possible sensor error, especially as ASUS AI Suite has an abysmally low refresh rate.