Your CPU has multiple temperature readings, and one might be making your fans scream

Published Oct 3, 2026, 2:30 PM EDT Ismar is a Senior Author at How-To Geek. He previously worked as a writer, editor, and general manager at a content agency before joining the team in 2023. He began his writing career in 2021 after completing a BA in English Language and Literature. Beyond How-To Geek, Ismar has contributed to Red Stag Fulfillment and Authority Hacker. He has also worked on various SEO projects to help clients improve their search visibility. Ismar has been around Windows PCs since the age of three, so friends and family naturally chose him as the resident tech support. His projects often involve tinkering, such as disassembling mechanical keyboards and gaming mice to mod his beloved gear. He also enjoys pushing hardware to the max through overclocking while staying on a reasonable budget. An avid gamer, Ismar has logged thousands of hours across various genres, including first-person shooters, RPGs, racing games, and roguelites. When he is not at his desk, he is probably exploring other people’s phones to discover their quirks and features or experimenting with the latest AI tools. In his free time, Ismar enjoys spending time with his wife, working out at the gym, playing guitar and bass, cooking gourmet meals, and traveling. Above all else, he’s a proud cat owner. It's easy to fall into the trap of thinking of hardware temperature as one-dimensional: one GPU temperature, one RAM temperature, and one CPU temperature. However, almost every one of those components can actually have several temperature readings spread across its tiny surface area, with the CPU often having the most complex temperature readings. Your CPU doesn't have just one temperature AMD and Intel measure CPU temperatures differently, though Credit: Will Verduzco / How-To GeekCredit: Ismar Hrnjicevic / How-To Geek Modern CPUs don't have just a single temperature reading; there are numerous microscopic thermometers that report temperatures across the silicon die and, more often than not, outside the main die too, such as the IOD (I/O die). These temperature readings can sometimes have slightly different names depending on the app you use to monitor them, so for the sake of simplicity, I'll just focus on the readings as they appear in HWiNFO throughout this article. When you open HWiNFO in sensors-only mode for the first time, you'll be met with a long list of sensors, readings, and metrics, but the CPU is what we want to focus on. Intel and AMD measure CPU temperatures slightly differently, so the readings aren't directly equivalent, but they provide broadly similar information. On AMD, you'll generally want to look at the CPU Model: Enhanced section, while on Intel, the most useful CPU temperature readings are typically found under CPU Model: DTS. AMD Ryzen exposes a couple of key CPU temperatures you should be aware of: CPU (Tctl/Tdie) and CPU Die (Average). The Tctl reading tells you the highest temperature among the relevant internal sensors. On some first-generation Ryzen chips, like the 1800X, Tctl also had an artificial +20°C offset, which could cause fans to ramp up unnecessarily. This is important for our discussion later, as Tctl is often what we think of as the “real” CPU temperature, even though it represents the hottest reading from a set of internal sensors. CPU Die (Average), on the other hand, is a short-term average of multiple temperature sensors on the CPU die. Unlike Tctl, which can fluctuate rapidly, this reading gives you a better sense of the CPU's thermal state under sustained loads. Depending on what you're doing, it can be considerably lower than Tctl. There are also per-core and per-CCD temperature readings, along with readings for the IOD and SoC, but these are generally most useful when you want to see which core or CCD is hottest or when troubleshooting a specific issue. On Intel, the most important CPU temperature readings are CPU Package and Core Max (which is different from TJMax). CPU Package represents a 256-millisecond average of the hottest temperature reported by the digital thermal sensors within the CPU package, making it somewhat similar to Tctl. For most people, this is the temperature that you should monitor, and it's what typically affects the fan speed on most motherboards. Core Max, on the other hand, tells you the hottest individual CPU core, making it useful for spotting core-level peaks. A large temperature gap doesn't mean your cooler is failing, but it's worth investigating The delta tells you something, but it's not always cause for concern Allow me to preface this by explaining that temperature gaps between the various CPU temperature readings are completely normal. Tctl on AMD and CPU Package on Intel can reflect the hottest part of the CPU, so they can be surprisingly spiky. They react to the hottest sensor reading, which can sometimes be the result of a quick, localized burst of activity, like opening an app or even moving your cursor around, before quickly dropping back down. For that reason, focusing on the average reading in HWiNFO gives you a more realistic representation of where the temperature typically hovers during your usual workloads, like gaming or video editing. On Intel, the temperature delta (difference) between CPU Package and Core Max doesn't matter much because they don't measure the same thing. Simply monitor the CPU Package and make sure it stays as cool as possible. On Ryzen chips, the delta between Tctl and CPU Die (Average) can be a little more informative. A temporary gap of several degrees in the current temperature isn't a big deal, but it's a slightly different story if that gap shows up in the Average column of HWiNFO during a heavy workload. If you run a heavy synthetic benchmark to push your CPU to its limit, like AIDA64, and notice a persistent gap after several minutes of sustained load, it's worth investigating your cooling setup, as there's a chance you've got dried-out thermal paste or uneven cooler pressure. I briefly ran a stress test on my Ryzen 7 7700, and there was virtually no gap between the two readings. During regular use, however, Tctl is on average about 4°C higher. Your fans may be reacting to every tiny temperature spike, but there's an easy fix Don't let those brief spikes control your fans One of the most annoying behaviors a desktop PC can exhibit is rapid fan step-up and step-down. One moment, your PC is completely silent; you open Chrome, and in an instant, the fans are louder than a fighter jet before dropping back down after only a few seconds. This can happen because the motherboard monitors a CPU temperature that fluctuates rapidly as the load changes, such as Tctl on AMD Ryzen. To stop this annoying fan behavior, head over to your BIOS and look for the step-up time option. A small delay of a second or two is often all it takes to prevent this behavior, but depending on how large your heatsink is, you can probably get away with a much more generous delay. For instance, my PC has a dual-tower air cooler, so it takes quite a while for the heatsink to soak up the heat from the CPU before the fans even have a chance to blow fresh air across it. I've set all my CPU and case fans to ramp up only after 25 seconds, combined with a fairly loose fan curve that gradually ramps the fans up from 70°C onwards. This means the fans only kick in when there's an actual sustained load on my CPU that warrants additional cooling. During regular desktop use, the fans remain almost inaudible at the 20% fan-speed baseline. CPU temperatures are a lot more complicated than they seem at first The next time you see your CPU temperature jump without warning, don't assume your processor instantly went from cool to overheating. Modern CPUs are packed with temperature sensors, and they can report their hottest readings, which the motherboard may then use to adjust fan speed. Those brief temperature spikes are a completely normal part of how a processor operates, and they're no reason to panic. Just adjust your fan curve so those annoying speed changes don't bother you when they're not needed.