Temperature is a complicated thing to figure out if it's "right" or not. So if you're sure you did the right things like make sure to remove the plastic covering the heat sink, made sure there was ample coverage of the thermal paste, made sure the screws were at least reasonably tight, then it's time to really look at what's going on.
For one, a modern processor reports multiple things as temperature. For example, here's what HWiNFO reports of my laptop's CPU, an i7-13700HX
There's 8 readings, what the heck am I supposed to be concerned about? Well thankfully HWiNFO has tool tips if you hover over one and it recommends paying attention to "CPU Package", which is a 256ms average of
all of the sensor readings. Note that sensors is what the processor has, it's more than what HWiNFO can pick up.
If the application you're using monitors only a single temperature, you have to figure out which exposed sensor it's reporting. One app I use, NZXT CAM, seems to use the "Core Temperatures" value, which is an average of all the core temperatures at that point when the reading was updated. This isn't as useful as the "CPU Package" temperature because it may report temperature spikes, which isn't really a cause of concern. Temperature over time is a much better indicator of overall cooling "health" than temperature for a split second. It's why you can "karate chop" a candle flame: your skin isn't touching the flame long enough for it to matter (among other things, but that's part of it)
So go run Cinebench R23 for the 10 minutes it does by default. If the CPU temperature only hits 100C some times, it's fine. If it's constantly near 100C, then we have a problem.
Facebook
Twitter
Instagram