How to Check CPU Temp on Windows 11 (and GPU Temp Too)

Windows has no built-in CPU temperature readout. The two reliable ways to see it are your motherboard’s BIOS screen, which needs nothing installed, and a free monitoring tool. Graphics card temperature is different: that one Windows does show you.

Last updated: September 2026

The Short Answer

• Nothing installed, right now: restart, enter the BIOS, and read the temperature on the first screen. Works on every desktop motherboard.
• Inside Windows: install Core Temp for just the processor, or HWiNFO if you want every sensor in the machine.
• While gaming: an on-screen overlay, because the temperature that matters is the one under load, not the one after you alt-tab.
• Graphics card: Task Manager, Performance tab, GPU. No download needed.
• The command line trick you may have read about: it no longer exists. The section near the end explains what replaced it.

Windows 11 Does Not Show CPU Temperature Anywhere

This is the part most guides skirt around. There is no CPU temperature in Task Manager, none in Settings, and none anywhere else in the operating system. Microsoft’s own Task Manager documentation describes the Performance tab as “Performance | Lists and graphs of performance data. Separate views are available for CPU performance, memory performance, and other performance categories (depending on the computer configuration).” (Microsoft), and the detail it lists is processor load, kernel and user time, and per-logical-processor activity. Temperature is not among it.

Two near misses are worth knowing about, because both send people down dead ends.

The first is Performance Monitor, which does have a Thermal Zone Information counter set. What that reports is an ACPI thermal zone: a region the system firmware declares, with one sensor, used by Windows to decide when to throttle. On most desktops it is not the processor die and on many it is not populated at all.

The second is the Windows Management Instrumentation class Win32_TemperatureProbe, which looks exactly like what you want until you read Microsoft’s note on it: current implementations of WMI do not populate the reading. The property exists; nothing fills it in.

So the operating system genuinely cannot tell you. That is why every answer below involves either firmware or a third-party tool.

Method 1: Read It in the BIOS, With Nothing Installed

Your motherboard has always known the temperature, and it shows you on the first screen of its setup utility. Restart the machine and press Delete or F2 as it boots. The temperature is on the landing screen, next to the fan speeds.

Brand	Where to look	Source
ASUS	UEFI BIOS Utility – EZ Mode (press F1 / Del at boot)	manual
GIGABYTE	BIOS Setup – Easy Mode, and the Smart Fan 6 screen (F6)	manual
MSI	Click BIOS – EZ Mode, and the Hardware Monitor page (F5)	manual
ASRock	UEFI Setup Utility – Main screen and the H/W Monitor tab	manual

ASUS goes as far as colouring it: its own support page describes a CPU temperature of 85 °C on the EZ Mode screen and notes that “if it is displayed in red, it means the temperature is too high” (ASUS).

The limitation is obvious once you think about it. A BIOS reading is an idle reading taken with the operating system not running, so it tells you almost nothing about how hot the processor gets while you are actually using it. It is the right tool for two jobs: confirming the sensor works when Windows will not boot, and checking a machine that is overheating badly enough to shut down before it reaches the desktop. For anything else, use a tool inside Windows.

Method 2: A Free Monitoring Tool

This is the practical answer for almost everyone. All of these are free for home use.

Tool	What it shows	Cost	Download
Core Temp	Per-core CPU temperature, read from each core’s own sensor	Free for personal use	alcpu.com
HWiNFO (HWiNFO64)	Everything: per-core and package CPU, GPU, drives, fans, voltages, power	Free for personal and non-commercial use	hwinfo.com
HWMonitor (CPUID)	CPU, GPU, motherboard, drives, fan speeds, voltages, in one flat list	Free (a paid Pro version exists)	cpuid.com
Libre Hardware Monitor	Same job, open source, no installer needed	Free, open source	GitHub
MSI Afterburner	Graphics card control plus an on-screen display you can see mid-game	Free	msi.com

Core Temp is the one to start with if you only care about the processor. It reads each core’s own digital thermal sensor and shows the cores in a small window, with the current, minimum and maximum for each. Install it, open it, and the numbers are there; there is nothing to configure. Its developer describes it as “a compact, no fuss, small footprint, yet powerful program to monitor processor temperature and other vital information”.

HWiNFO is the one to use when you want the whole picture, because it reads the graphics card, the drives, the fans and the voltage rails as well. Run it in Sensors-only mode and you get a long list; the entries you want are CPU Package and the per-core rows. HWiNFO’s licence says “All versions are free to use in personal and non-commercial environments.”, with a paid version only for business use.

Method 3: The Tool From Intel or AMD

Both processor makers publish their own utility, and both come with a catch worth knowing before you download.

Intel Extreme Tuning Utility. Intel describes it as a tool to “The Intel® Extreme Tuning Utility (Intel® XTU) is used to overclock, monitor, and stress a system.” and it does show package temperature. The catch is in Intel’s own support note: “This application is supported with unlocked processors (K- and X-series) only.” If your processor is not a K or X model, and your board is not a Z-series, this is not your tool (Intel).

AMD Ryzen Master. AMD’s utility monitors per-core clock rates, temperature and voltages, including average and peak readings, and it works across the Ryzen range rather than on unlocked chips only. It is the right choice on a Ryzen system for one specific reason covered in the sensor section below: it is the tool that reports the true die temperature rather than the offset figure other software may show (AMD).

Method 4: Watch It While You Are Gaming

The temperature that matters is the one during the workload, and alt-tabbing out of a game to look at a window changes the thing you are measuring. An on-screen overlay solves that.

MSI Afterburner is the usual answer, and MSI is explicit that it “works on any card (even non-MSI!)”. Install it with the RivaTuner component it offers, open the monitoring settings, tick the values you want and enable “show in On-Screen Display” for each. The numbers then sit in the corner of the screen while you play.

One accuracy note: MSI’s own product page describes Afterburner as graphics card software with hardware monitoring, and does not claim it reports processor temperature. On most systems it does, through the same sensor layer other tools use, but if the row is missing for your board, run HWiNFO alongside it rather than assuming the reading is unavailable.

How to Check GPU Temperature: Windows Does This One For You

The graphics card is the exception. Open Task Manager with Ctrl+Shift+Esc, go to the Performance tab, and select GPU. The temperature is next to the graphs.

Microsoft added it in a Windows Insider build in 2019 and published the conditions, which still explain most cases where the number is missing:

• “Showing the GPU temperature in Task Manager is currently only supported if you have a dedicated GPU card.”
• “You may need an updated graphics driver to see the temperature. Specifically one which supports version 2.4 (or higher) of WDDM is required. To check which version your PC has, press WIN+R and enter dxdiag. Under the Display tab, it will say your WDDM version next to Driver Model.”
• “Currently the temperature value is only supported in Celsius.”

(Microsoft, Windows Insider Blog) So: a dedicated graphics card, a driver new enough to use WDDM 2.4 or later, and Celsius only. On a laptop running on integrated graphics there is no reading to show, and on a very old driver the row simply does not appear. Update the driver first, then check again.

For anything more than the current number, the same tools from the section above report the graphics card’s core, hotspot and memory temperatures. Our guide to normal GPU temperatures has measured figures for individual cards to compare yours against.

Why Two Tools Show Two Different Numbers

Open Core Temp and HWiNFO at the same time and you may see readings a degree or two apart, or labels that do not match at all. That is not a bug in either one. It comes from how processors report temperature in the first place.

On Intel: the sensor counts backwards

An Intel processor does not store a temperature. Its digital thermal sensor stores a distance from the throttling point. Intel’s datasheet puts it exactly: “The temperature returned by the DTS is an implied negative integer indicating the relative offset from TjMAX. The DTS does not report temperatures greater than TjMAX.” and “It is the responsibility of software to convert the relative temperature to an absolute temperature. The absolute reference temperature is readable in the TEMPERATURE_TARGET (0x1A2) MSR.” (Intel).

So every tool has to do the same sum, from a reference value it reads separately, and a tool that gets that reference wrong is wrong by a constant offset. Two more things follow from the same datasheet. The sensor has a stated accuracy of “±5 °C within the entire operating range”, and it “does not report temperatures greater than TjMAX”, which is why a processor pinned at its limit shows a suspiciously steady number.

The point it counts down to is Tjunction max, which Intel defines as “Tjunction max is the maximum thermal junction temperature that a processor will allow prior to using internal thermal control mechanisms to reduce power and limit temperature.” (Intel). You will also see Tcase quoted in specification tables. That is a different thing entirely: the maximum allowed at the metal lid, measured in a lab with an embedded thermocouple. No software on your PC can read it, so a tool showing you a live number is never showing you Tcase.

On AMD: the number may be deliberately inflated

AMD’s main reporting sensor is tCTL, and AMD has explained that it is not always the real temperature: “The tCTL sensor is derived from the junction (Tj) temperature… but it may be offset on certain CPU models so that all models on the AM4 Platform have the same maximum tCTL value. This approach ensures that all AMD Ryzen processors have a consistent fan policy.” (AMD).

The clearest documented case is the Ryzen 7 1700X and 1800X, which AMD says “carry a +20°C offset between the tCTL (reported) temperature and the actual Tj temperature”. AMD’s fix was in software rather than in the sensor: Ryzen Master was updated to report junction temperature by removing the offset automatically.

Practically: on a Ryzen system, the row labelled Tdie or CPU (Tdie) is the real die temperature and the one to trust. Tctl is the fan-control figure. If a tool shows only one number and it looks high for the workload, check it against Ryzen Master before worrying.

And a reading that jumps around is normal

Intel documents two ways of reading the same sensor: an instantaneous read, and one averaged over a 256 millisecond window intended for fan control. That is why the BIOS shows a calm number while software shows brief spikes to figures you never see in the firmware. Both are correct; they are answering slightly different questions. Judge your cooling on where the temperature settles under sustained load, not on the highest value that ever flashed past.

The Command Line Trick That No Longer Works

A one-line command has circulated for years as the way to read CPU temperature without installing anything:

wmic /namespace:\\root\wmi PATH MSAcpi_ThermalZoneTemperature get CurrentTemperature

There are three separate reasons not to rely on it, and the first is decisive. Microsoft has removed the tool: its deprecated-features list now states “WMIC is removed and is no longer available as a Feature on Demand (FoD).” on Windows 11 version 24H2 and above (Microsoft).

Second, even where it still runs, the class reports an ACPI thermal zone rather than the processor. That is a region the firmware declares, and on most desktop boards it is either absent or measuring something other than the CPU die, which is why the command so often returns nothing or an implausible value.

Third, the number comes back in tenths of a kelvin, so it needs converting before it means anything.

The modern equivalent, where the class exists at all, is a PowerShell query run as administrator:

Get-CimInstance -Namespace root/wmi -ClassName MSAcpi_ThermalZoneTemperature

It has the same second and third problems. On a desktop, install Core Temp instead and spend the thirty seconds better.

Now That You Have a Number

A reading on its own does not tell you much. What matters is where it settles under load and whether that figure is normal for your processor.

• What is a good CPU temp has the ranges by workload, measured figures for current Intel and AMD chips, and each one’s published limit.
• Normal GPU temp does the same for graphics cards, per model.
• If the number is high, CPU overheating signs and fixes covers what to do about it, and the thermal paste guide covers when a repaste is actually worth it.
• If the fans are loud rather than the temperature high, why your PC fan is so loud is the better starting point.

Frequently Asked Questions

How do I check my CPU temp on Windows 11?

Windows 11 has no built-in CPU temperature readout, so you need either the BIOS or a free tool. For a reading right now with nothing installed, restart, press Delete or F2 to enter the BIOS, and read the temperature on the first screen. For a reading inside Windows, install Core Temp for the processor alone, or HWiNFO if you want the graphics card, drives and fans too. Both are free for personal use. Take the reading after ten to fifteen minutes under load, not at idle, because the idle figure tells you almost nothing about your cooling.

Can I see CPU temperature in Task Manager?

No. Task Manager shows processor load, not temperature, and Microsoft’s own documentation for the Performance tab lists no temperature field for the CPU. It does show a GPU temperature, which Microsoft added in 2019, but there has never been a CPU equivalent. If a guide tells you to look for it in Task Manager, it is describing the GPU tab.

How do I check CPU temp without downloading anything?

Use the BIOS. Restart the machine, press Delete or F2 while it boots, and the temperature is on the landing screen next to the fan speeds, on ASUS, MSI, Gigabyte and ASRock boards alike. The old command line trick using wmic no longer works: Microsoft removed the wmic tool from Windows 11 24H2 and later, and even where it ran it reported a firmware thermal zone rather than the processor.

Why does my CPU temperature differ between two programs?

Because the processor does not report a temperature directly. On Intel, the sensor stores the distance to the throttling point and each tool converts that to degrees itself, from a reference value it reads separately; Intel also gives the sensor an accuracy of plus or minus 5 degrees. On AMD, the reported tCTL figure is deliberately offset on some models so that every chip on the platform hits the same maximum and fan curves behave the same way, while the Tdie row is the real die temperature. A degree or two of disagreement is normal. A twenty-degree gap on an older Ryzen chip is the documented offset, not a fault.

Is Tctl or Tdie the real temperature on a Ryzen CPU?

Tdie. AMD describes tCTL as derived from the junction temperature but offset on certain models to give the whole platform a consistent fan policy. Tdie is that offset removed. AMD’s Ryzen Master removes it automatically, so if two tools disagree on a Ryzen system, Ryzen Master is the tie-breaker.

How do I check GPU temperature?

Task Manager does this one. Press Ctrl+Shift+Esc, open the Performance tab and select GPU; the temperature sits next to the graphs. Microsoft’s conditions are that you need a dedicated graphics card, a driver supporting WDDM 2.4 or later, and that the value is shown in Celsius only. On integrated graphics there is nothing to show. For hotspot and memory temperatures rather than just the core, use HWiNFO or your card vendor’s own software.

The Bottom Line

Two tools cover every case. The BIOS gives you a reading with nothing installed and works even when Windows will not start. Core Temp or HWiNFO gives you the reading that actually matters, the one under sustained load. Task Manager handles the graphics card on its own. Everything else, including the command line one-liner that no longer exists, is a detour. Once you have the number, the question becomes whether it is normal, and that is a different guide.