New 5600x temperature question

Joe Q

Member
I upgraded from a 3700x to a 5600c and was wondering if something was different with how hwinfo reports the temperatures? I’m asking because on my 3700x, tctl/tdie was within 5 degrees of tdie but also was always the highest value. With my 5600x, many times during mild load the tdie will report a higher temperature with a difference of up to 15C from tctl/tdie. After a two hour game session, max tctl/tdie was 59C but tdie max was 73C. Seems like a big gap
 
Temperature calculation is a quite complex process as there are several factors and values from dozens of internal sensors involved in the final result.
HWiNFO reports values calculated by the CPU internally, but we don't have all details how exactly these are calculated internally.
So I'm afraid I can't help here much, this should be better answered by AMD.
 
Hello I had a Ryzen 1700 and tctl/tdie was very close to the other so I took that as a measurement. I've upgraded to ryzen 5900X now tctl/tdie is off from CPU Die (avarege) like 9-10 degrees. In ryzen master the temperature is the same as the CPU Die (avarege) . I am using MSI B550 MEG Tomahawk board in bios the fan curve you can set it to CPU CORE or CPU Socket the CPU socket is what actually the temperature in Ryzen master. I am using EK Fluid gaming A240G and I have top mounted radiator on my case so I can easily feel the air temp in Cyberpunk 2077 when it says the cpu is doing 68-72c It's no where close to what the temp was when my old CPU was doing 65c (didn't get hotter than that). So for now I am to believe the CPU Die (avarege) and I set my fan curve based on the CPU Socket in MSI bios. One more friend with 5800x reports the same finidings that the tctl/tdie are way off from ryzen master temps which should be the most accurate ones.
 
Hello I had a Ryzen 1700 and tctl/tdie was very close to the other so I took that as a measurement. I've upgraded to ryzen 5900X now tctl/tdie is off from CPU Die (avarege) like 9-10 degrees. In ryzen master the temperature is the same as the CPU Die (avarege) . I am using MSI B550 MEG Tomahawk board in bios the fan curve you can set it to CPU CORE or CPU Socket the CPU socket is what actually the temperature in Ryzen master. I am using EK Fluid gaming A240G and I have top mounted radiator on my case so I can easily feel the air temp in Cyberpunk 2077 when it says the cpu is doing 68-72c It's no where close to what the temp was when my old CPU was doing 65c (didn't get hotter than that). So for now I am to believe the CPU Die (avarege) and I set my fan curve based on the CPU Socket in MSI bios. One more friend with 5800x reports the same finidings that the tctl/tdie are way off from ryzen master temps which should be the most accurate ones.
Thanks for your observation and feedback !
 
Just to add after some more testing it seems that the two readings tctl/tdie and CPU Die (avarage) have these 9-10 degress difference only on idle and light workloads like browsing when a heavy workload like cinebench or even a game is introduced the two readings are identical.
 
Thanks for the insight. I sort of forgot about this because I hadn’t really noticed it happening anymore. I did update to the latest version, so maybe that fixed it up.
 
Hello I had a Ryzen 1700 and tctl/tdie was very close to the other so I took that as a measurement. I've upgraded to ryzen 5900X now tctl/tdie is off from CPU Die (avarege) like 9-10 degrees. In ryzen master the temperature is the same as the CPU Die (avarege) . I am using MSI B550 MEG Tomahawk board in bios the fan curve you can set it to CPU CORE or CPU Socket the CPU socket is what actually the temperature in Ryzen master. I am using EK Fluid gaming A240G and I have top mounted radiator on my case so I can easily feel the air temp in Cyberpunk 2077 when it says the cpu is doing 68-72c It's no where close to what the temp was when my old CPU was doing 65c (didn't get hotter than that). So for now I am to believe the CPU Die (avarege) and I set my fan curve based on the CPU Socket in MSI bios. One more friend with 5800x reports the same finidings that the tctl/tdie are way off from ryzen master temps which should be the most accurate ones.
I have a 5800x and experience the same behavior. Honestly, I'm not quite sure which one to follow as well but since Ryzen Master reads out the same temp as CPU Die (average), this looks to be what should be monitored. Good thing for me too as this is usually the lowest temp amongst the three.
 
If you search here in forum you will find plenty explanations about temp readings on ZEN2/3. Apparently you haven’t...

“Tctl/Tdie” is the hotspot always reporting instantly (switching to) the hottest reading sensor.
“Tdie average” is the average of all sensors in package (? about IO die)
CCD1/2 is the edge temp of each CCD from a fixed location.

Idle or light loads makes the 3/4 readings to be different as different parts of the cores have different load.
Medium/Reduced threaded loads like gaming with make them closer to each other. Still Tctl should be higher.
Max load will make them report almost the same as heat is distributed almost equally on all die area.
 
I own also a 5900X and was wondering also which temp to rely on in HWInfo for the CPU.
I read a lot of topics (also in HWinfo of course), and I still don't know which one we should consider as closest to Ryzen master's one : CPU Die (Average) or CPU CCD1 (Tdie) ? (Both temps are the closest one's to Ryzen master reported temperature for me).

Thank you in advance.
 
If you search here in forum you will find plenty explanations about temp readings on ZEN2/3. Apparently you haven’t...

“Tctl/Tdie” is the hotspot always reporting instantly (switching to) the hottest reading sensor.
“Tdie average” is the average of all sensors in package (? about IO die)
CCD1/2 is the edge temp of each CCD from a fixed location.

Idle or light loads makes the 3/4 readings to be different as different parts of the cores have different load.
Medium/Reduced threaded loads like gaming with make them closer to each other. Still Tctl should be higher.
Max load will make them report almost the same as heat is distributed almost equally on all die area.
Hi, thanks. Tdie for my 5600x shows 69.0 max after hours of gaming, Tdie average shows 109.9! Ccd2 shows 73.3. Any idea what is happening?
 
Last edited:
I am experiencing the same issue with my Ryzen 5 5600x

Average temperature is way above reported max temperature.

1645826069556.png

But you can see in the screenshot that the processor never reported thermal throttling, which it definitely would have if it ever actually reached 109 C ...

And sorry but I don't use Ryzen Master so I can't compare what it says.
 
No; not even sure what either of those things are.

From googling what Hydra is, I assume they're both overclocking tools. I'm not running any kind of overclock on CPU or GPU. Just running stock 5600x and stock RX 6700 XT.
 
That's certainly a wrong readout. Looks like something else in the system is interfering with HWiNFO causing this.
 
No; not even sure what either of those things are.

From googling what Hydra is, I assume they're both overclocking tools. I'm not running any kind of overclock on CPU or GPU. Just running stock 5600x and stock RX 6700 XT.
What software do you have configured to start with Windows? You can see the items by opening Task Manager -> Startup, or Settings -> Apps -> Startup (Windows 11). You might also find items in "C:\Users\YourUsername\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup" - replace YourUsername with your actual username.
 
What software do you have configured to start with Windows? You can see the items by opening Task Manager -> Startup, or Settings -> Apps -> Startup (Windows 11). You might also find items in "C:\Users\YourUsername\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup" - replace YourUsername with your actual username.
I don't keep a lot of stuff running in the background. My PC is self-built so there's not any bloatware on it, and I keep it pretty lean.

1646519208819.png

Since the only thing there you might not recognize is vol_control.exe, I'll explain that it's just a self-written Autohotkey script that lets me press different key combos to perform media controls on a plain old keyboard that doesn't come with media buttons.
 
I don't keep a lot of stuff running in the background. My PC is self-built so there's not any bloatware on it, and I keep it pretty lean.

View attachment 7478

Since the only thing there you might not recognize is vol_control.exe, I'll explain that it's just a self-written Autohotkey script that lets me press different key combos to perform media controls on a plain old keyboard that doesn't come with media buttons.
Can you easily reproduce the erroneous temperature reporting, or is it only under certain conditions it appears?
 
Back
Top