So i updated the XTU in hope it will stop crashing when the PC runs for longer than 4 days, just to see this here. Ain't that a gem!
I have seen such spikes quite few times, but i did not care to check the value, so this time i actually did.
Win10, XTU ver. 7.14.015, i7 13700k. Motherboard Gigabyte B760 Gaming X DDR4.
My question is, am i the only one seeing this? This is definitely a bug.
I have not tried to run the XTU without any TDP limit, but i will definitely do that, to see if the bug occurs only when TDP is limited.
Link Copied
- May we know the exact issue (Please share more details about the issue so we can assist you effectively)?
The exact issue is incorrect readings/data shown by the XTU software.
Since my PC is water cooled, i rely the graph to know if the water cooling is working as it supposed to do.
Now, i would end it here, but since i posted on here, there is another odd thing:
'Power limit throttling' jumps on, i even went with teh mouse over, to see if it's just the graph, but no, it also says on the very left 'Yes', so while there is a 125 Watt limit on the CPU, there is no spike in the temperature nor power reading:
By now i can easily reproduce the spikes on the 'Power limit throttling' , for that i just need to run the Google Chrome browser.
Or i can launch Egde browser, and just by closing it, i can reproduce such spikes each time!
So, Chrome does the spikes randomly, while closing Edge results in a spike each time.
However, i am unable to reproduce the abnormal value on the 'Package TDP' reading, just the Power limit throttling for now.
Firefox browser does not produce such spikes at all.
- Is the system working before (no crashes, no bugs)?
If XTU was no launched nor running in the backgroung, then the PC would not crash up to 18 days, that is the max i was willing to test it.
With XTU running, the PC will crash within few days.
So there is a connection to XTU running on my PC.
- Any recent changes made to the system?
Nope, no Hardware nor software changes except the XTU. In fact, it will be close to 1year my PC gets no windows updates.
Afterdisabling all WIndows updates my PC runs as smooth as never!
- Have you overclock the processor?
The motherboard i use is the B760 from Gigabyte, it does not allow overclocking. And since i did not perform any changes in BIOS for maybe longer than 2years, i see no connection here.
Summary, previous XTU version did never show such spikes, this one does.
SSU Log shows one reading that can't be correct, under 'Processor' it says "Current Voltage:1.7 volts" which is literally not possibe, as about 1,43V is the max it can go.
And this part of the XTU shows exactly that:
Thanks for the reply, on my side it is not a real problem, but more of an inconvenience.
And obviously, i just selected only few readings for the graph, maybe there is more wrong readings, but there is like 100 of them, or more, so i can't possible check all of them.
Hmm,
so i don't have CPU Z installed, nor any related services, but i have Corsair RAM and iCUE software installed, along with the corsair service(s) running in the background.
Never before i had any false readings (at least no spikes) until i upgraded the XTU to Version 7.14.015, i am not sure what version i had before.
So regardless what was running, there were no spikes on the previous version of XTU.
With no browser open, there will be no spikes visible on the XTU graph - even if iCUE running along with all Corsair service(s).
The moment i open Chrome or Edge = a lot of spikes.
Firefox does not produce spikes at all......
I just uninstalled everything Corsair related and restarted the PC - i can still reproduce the spikes.
And before i uninstalled Corsair software, Chrome would randomly create those spikes, now it seems it does not, only when closed or opened.
In screenshot - first spike came after opening Chrome, the second one when i closed Edge, closing Edge results almost each time with a spike.
As can be seen, now i have no services related to Corsair.
However, there are definitely Corsair drivers on my PC - bcoz the RAM is from Corsair.
There is still that Gigabyte Control Center, which also monitors frequency and temperature on the CPU and other componnets, however, i can not uninstall the serivces, as i need it to control the fans, and some services are even required.
Again, i had it always running, so, i can't see it being related to the spikes either.
Anyway, thanks for looking into this matter.
Thanks for the reply.
- Could you please tell us if the crash happened on XTU or PC?
PC had a BSOD.
I leave the PC running 24/7, sometimes for many days or even weeks. Now i tested it with not just XTU but also other software to find out which one is causing the BSOD. After a very long period of time, i was able to pin it down to XTU.
Since with XTU running the PC would crash after few days, without XTU the PC would run for longer than 2 Weeks, and i tried this few times.
As you can imagine, this took months to perform, it is not something i can do frequently. - Do you see any error messages before the crash?
I checked the Evemt Viewer, but there was no other events than 'Kernel power' , so it's either Software/Hardware, i never witnessed the crash while using the PC, i use the PC only for few hours, and it's more like 18~20 hours when the PC runs unsupervised. - What are the symptoms of the crash? If you are only using XTU for monitoring, you can try to restart the program. Please note that XTU does not support very low program settings.
I can not name any symptoms, except that due to testing i found out that with XTU running, the PC will be having a BSOD. - You mentioned that the XTU software showed incorrect readings/data. What kind of data was incorrect except for the information that you already provided? Some spikes are expected (depending on your activity and background programs), as long as they return to normal after a while.
So it is very easy to notice spikes that make no sense if it's a value as ridiculous as in the screenshot, secondly i would not know if the value is incorrect for data such as frequency, i can not know if one core is in fact ruiing with such frequency or not.Therefore readings such as temperature, TDP or Power Limit Throttling are somethign i can tell if it makes sense or not.
I have no idea what has changed, with the limit i can replicate the spikes each time, but without the limit i can no longer do that.
Just now i tried to replicate the TDP limit spikes, but i can no longer replicate them, if i set NO TDP limit.
So, TDP limit ON (in this case 125Watts) = TDP limit spikes
No TDP limit (TDP set to 'Default' profile, this means the CPU may draw as much as 285Watts) = no TDP spikes
- Did you update the driver to the latest version?
Yes, all drivers are up to date. - You reported that the power limit throttling turns on and said that you can cause spikes on the power limit throttling by using Chrome or opening and closing Edge. Also, the spikes are consistent. However, Chrome causes random spikes, while closing Edge causes a spike every time.
Not exactly, it shows that TDP limit is on, but it's just a spike, just like on one of the screens a TDP value of about 16 digits number (few Terra Watts power usage..... hahahaha ) is impossible - which means, the graph shows something, which may be only a flase reading.
And like i stated above, i no longer can replicate those spikes with default TDP/unlimited. - What is the problem with these spikes? Spikes are normal when using different programs, as long as they are temporary and the readings go back to normal.
I never had such spikes, neither on TPD limit nor anything else, so for me such spikes that go to maximum are not normal. - You also asked why the graph kept moving even when paused. This happens because you changed the settings for system graphs, which probably restarted the run time (since you have a new selection). We have notified the XTU team about this and they will work on fixing it in the future.
Yes, this is definitely a bug, as the button that shows the state of logging is showing 'Pause' while the graph continues to show new readings.
And now i can do something i could not do before, the graph is completely broken as part of it continues to run, and part of it does not.
I am not sure what i happening right now.
This is what happened now after i tried to replicate the false state of the 'Pause' button:
CPU utilization shown in graph is still on Pause, but the rest is not, so the CPU utilization actually does what the Pause button shows, but the rest of the readings is just progressing.
I recorded this on video -> https://youtu.be/Tr520NQRkuI
So i recon that there are different bugs when on default TDP and when TDP is limited? i can not claim it was like this before, at least i could not make the graph pause part of the readings while continuing to show other readings in real time.
• Are you only using XTU for monitoring purposes?
Yes, monitoring, and to limit the CPU power with the Power Throttling, my water cooling is capable of running 24/7 with the CPU at 230W, but also with a temperature close to +100°C , obviously this is not my target.
• What is your goal of using XTU?
Monitoring.
• Could you please send us a screenshot and a log file: XtuCore logs available in C:\ProgramDate\Intel\Intel Extreme Tuning Utility\Logs..
The log files are quite big in dimensions, so i can not post a screenshot, buz i can upload few of the logs i have there, seems those extend only to the last update of the XTU which was at the end of January this year. So i don't have any older logs from the time the BSOD happened.
can't edit my last post, so i will add this one last reply.
I am able to reproduce the PC crashing almost each time - if i run AIDA64 'Julia' test while the TDP is not limited, and all BIOS settings are at max; XTU stress test does not crash the PC on same settings.
Each time AIDA64 crashed the PC, the PC would just like that restart without loading the default BIOS settings....so i guess the BIOS did not detect any settings failure.
And it is not really a BSOD, bcoz there is no message on the monitor, nor a dump file, the only thing that has some information is the Event Viewer, while the AIDA64 shows high temperature:
2nd time i set TDP to 250W with XTU, and the PC did not die on me,
3rd try i set the TDP to 290W and again no death, but i noticed, that this test shows up in XTU as a TDP spike that goes sharp to the max set TDP of exactly 290W, so unlike XTU, 'Julia' test spikes each time to max TDP
4th time TDP was set to 125W and the PC died again, the indication of the TDP is the temperature at close to only +50°C:
What AIDA64 did with that test, resembles what other users here expirience with their PCs, except, i need to run that test and i don't get any BSOD, the PC just dies and restarts as if nothing happend.
And without that i have no problems with the PC at all.
That message on red background appears only if i start the stability test, then stop it, and then i run that Julia test, then the PC crashes and restarts, and then i need to run AIDA64 and click on the stability test without starting it, so i see it after the PC 'recovers' from the crash.
