Flying over western Europe do you find recurrent spikes...

[Cpt_Piett 4,833]

I never reported back after finishing the LIRF-EHAM flight last night. I did collect the data though. Used AutoFPS with TLOD min 50 below 1000ft and max 975 (I know, it’s very high… I can explain… lol).

Core temperatures from FL200 climbing out from LIRF until end of session:

Approach (AIG traffic controller set at 50%):

Edited May 4 by Cpt_Piett

[hvw 416]

12 hours ago, YVR_23 said:

The EU is PG heavy so if you have your settings on max you're going to see spikes. One thing that's helped me is going from ultra to high clouds. This takes the load off my GPU and keeps in the temps down.

PG is off here, too. and clouds are set to high. With TLOD at 75 and OLOD at 50 I do get very reasonable temps. Right now I'm on my way from EDDK to LEZL and my CPU temperature is 63-66C GPU at a constant 47C, Not bad, uh? (All WU's and CU's are deleted and I'm not using MSFS_AutoFPS).

[Noel 4,563]

21 hours ago, YVR_23 said:

The EU is PG heavy so if you have your settings on max you're going to see spikes. One thing that's helped me is going from ultra to high clouds. This takes the load off my GPU and keeps in the temps down.

I had PG off when I made the original post above, but I've turned it back on after someone stated PG on/off makes no difference with this issue.  These "spikes" were every 10 seconds on average and nailed the main thread and with it spikes in CPU temp in fact one to 90C and several to 89C with fan maxing out.   GPU is usually asleep when this happens.  When this happened to me previously on a 9900K it was the same thing--you could be at cruise flying over what amounts to nothing (rural, mountains, etc) and it's tied to very specific regions it would appear.

I did LIRF > LSZH this morning and had no significant temp spikes beyond one to 82C but most of the time around 66-68C the whole flight.   TLOD I left at 100.  So at this point....

1. Uninstall/Reinstall of WU's helped somehow.  Of note again:  not had a recurrence of this problem since doing this, and my install was super fresh.
2. The exact route makes a difference so perhaps I dodged all the bad zones in the several flights I've done since reinstalling all WUs.
3. City Updates might play a big role in this as I've not reinstalled these.

 

[hvw 416]

25 minutes ago, Noel said:

..

1. Uninstall/Reinstall of WU's helped somehow.  Of note again:  not had a recurrence of this problem since doing this, and my install was super fresh.
2. The exact route makes a difference so perhaps I dodged all the bad zones in the several flights I've done since reinstalling all WUs.
3. City Updates might play a big role in this as I've not reinstalled these.

Thanks for sharing your findings. Tomorrow I will reinstall WU4 and WU6 again and see what happens here. I will, though, keep my TLOD at 75 and OLOD at 50 as these settings gave me the best results. And I don't mind that I "loose" 25 kilometers in detailed sight, I am at high flight levels anyhow. Like you, I won't install CU's for now

[iborg1979 101]

For that matter, I did LTFJ - EBLG today and performance indeed became... atrocious may be too strong a word, but definitely "not normal" as I reached the Germany-Belgium border area. Despite being at cruise, in the PMDG 737 which is not the worst addon in terms of performance, as well as AutoFPS.

On final approach it reached near slideshow state. 

And now I'm flying over the USA and it's ultra-smooth.

[Noel 4,563]

2 hours ago, hvw said:

Thanks for sharing your findings. Tomorrow I will reinstall WU4 and WU6 again and see what happens here. I will, though, keep my TLOD at 75 and OLOD at 50 as these settings gave me the best results. And I don't mind that I "loose" 25 kilometers in detailed sight, I am at high flight levels anyhow. Like you, I won't install CU's for now

Some new but consistent findings:  now doing EHAM LIRF and all was great until I was getting close to flying over Dusseldorf at which point started spiked main thread and temps to 87C or 89C then back down again.  TLOD was left on 100.  Then in the midst of these started DynLOD which reduced down TLOD to 30.  This ended these spikes.  Shortly thereafter, which was about 5 minutes ago we were flying over EDDF and no spikes here, w/ TLOD at 30. Fortunately for THIS flight there is dense overcast so can't seem the poor quality TLOD.  But....perhaps one can decide WHEN they will be doing these flights over parts of Germany/Switzerland as apparently these are the worst of them...in other words aim for lots of weather!  Or, as our friend Ixoye suggests just don the headphones and don't let 89C bother you...and I think that is sage advise as you're certainly not going to hurt 7800X3D with that temps especially as transient spikes, which is really all they are.

 

[hvw 416]

8 hours ago, Noel said:

Or, as our friend Ixoye suggests just don the headphones and don't let 89C bother you...and I think that is sage advise as you're certainly not going to hurt 7800X3D with that temps especially as transient spikes, which is really all they are.

He is right, but doing so gives me an unsatisfactory feeling that once all was well and why is it that we cannot get it back to this old situation? It itches, so to speak. But again, he is right. Especially when you read Intel website about CPU temperatures, They state that high temperatures are not ideal, but that in fact nothing can happen to the CPU when Tjmax has bean reached. In such a case the CPU will throttle back and we just can go on with what we were doing. Gives me a strange feeling, too 🙂 I won't take the risk to find out if indeed Intel's statement is correct.

Perhaps I should give DynLOD a try as well, later today I will go and try to find where to obtain it.

[Noel 4,563]

3 hours ago, hvw said:

He is right, but doing so gives me an unsatisfactory feeling that once all was well and why is it that we cannot get it back to this old situation? It itches, so to speak. But again, he is right. Especially when you read Intel website about CPU temperatures, They state that high temperatures are not ideal, but that in fact nothing can happen to the CPU when Tjmax has bean reached. In such a case the CPU will throttle back and we just can go on with what we were doing. Gives me a strange feeling, too 🙂 I won't take the risk to find out if indeed Intel's statement is correct.

Perhaps I should give DynLOD a try as well, later today I will go and try to find where to obtain it.

I have that unsatisfactory feeling too!  But yes, it's only for a brief second so not like it's sitting there all day at TJmax.   My plan over Europe is to run w/o DynLOD and launch it only if I run into these issues, which so far has really only been the one flight described. 

[hvw 416]

49 minutes ago, Noel said:

I have that unsatisfactory feeling too!  But yes, it's only for a brief second so not like it's sitting there all day at TJmax.   My plan over Europe is to run w/o DynLOD and launch it only if I run into these issues, which so far has really only been the one flight described. 

This morning I decided to install WU4 and WU6 again and do another LEZL-EDDK flight. Wished I wouldn't have reinstalled these two packs, as I was right back from where we started 🙂 Many high temperature temps, up to 95C, As soon as I finished this flight first thing I did was deleting these two packs again. Are we running around in circles, I wonder? I understood from some messages in this forum that this week the much awaited SU15 will be released that will bring as many great updates a.o. better memory management. Perhaps with this SU15 our problem will be solved, although my hopes are not all that high.

I guess we have to come to a standstill for the time being until we have seen if SU15 indeed brings the resolution to this here undocumented feature and take it up from there. What do you say?

[Ixoye 1,841]

I have set TJmax at 85C so my CPU never exceeds that and the performance is the same, It normally lies around 60-62C with short spikes to max 85C at the problem areas.

[tfm 196]

2 hours ago, Noel said:

I have that unsatisfactory feeling too!  But yes, it's only for a brief second so not like it's sitting there all day at TJmax.   My plan over Europe is to run w/o DynLOD and launch it only if I run into these issues, which so far has really only been the one flight described. 

The newest ones (14900k+) are spec’d to run at 100C.

Intel tech support have confirmed that this is a safe temperature to run at 24/7. The “problem” is that throttling kicks in to stop the temperature going ABOVE 100% - and this takes so much energy out of the CPU that it drops straight back down to a lower temp within seconds, so you’ll never see it running at 100C more than fleetingly .

[hvw 416]

12 minutes ago, Ixoye said:

I have set TJmax at 85C so my CPU never exceeds that and the performance is the same, It normally lies around 60-62C with short spikes to max 85C at the problem areas.

 

12 minutes ago, Ixoye said:

It normally lies around 60-62C with short spikes

That's what I had before the issue started way back then. And now I can achieve this result by uninstalling WU4 and WU6 and setting TLOD to 75. I think that when I set TJmax to 85 and reinstall WU4 and W6 again, I still will get high (higher than 68C) CPU temperatures. So for now, after lots of discussion in this thread, I will stick to the aforementioned settings.