Overclock.net banner
Cancel
Create thread New Forums

NEW! ClockTuner for Ryzen (CTR) 2.1 RC5

1 reading
Jump to Latest
741 - 760 of 847 Posts
#741 ·
Seishin said:
Reporting that version 2.1 rc5 does not work for me. I´m not using the auto tuning, just manually input frequency and save profile 1. It loads at windows start but takes no effect, if I deactivate profile and reactivate it takes effect sometimes, sometimes needs two rounds of deactivation-activation. Everything is up to date and in order with my PC. I´m using now version ctr 2.0 rc5 and it work as intended.
Click to expand...
I have the exact same problem. I just updated to the 2.1 RC5 version and the profile says it's being applied. I check and the processor is sitting at 3.6. I have to open the profile management, deactivate and activate for the profile to actually be applied.

Anyone else having this problem?
 
#742 · (Edited)
Syjeklye said:
I have the exact same problem. I just updated to the 2.1 RC5 version and the profile says it's being applied. I check and the processor is sitting at 3.6. I have to open the profile management, deactivate and activate for the profile to actually be applied.

Anyone else having this problem?
Click to expand...
In what situation are you looking at the frequency? Those profiles do NOT set a constant frequency. The CPU must be at or above the load % that is specified in that profile, eg. over 25% or over 80% for it to be activated.
Try loading cinebench after system startup and see which profile is active. The active one will have the rocket symbol in CTR. But also only if the CPU load is above its threshold
 
#743 ·
St54Kevin said:
Well the 5900X is a completely different CPU, so you cannot compare those two esepcially thermal wise. And BTW CPUs are designed to work falwlessly at their max temp, in this case 90C, so 80-82 is definitely not pretty hot.
Nevertheless I set Platform thermal max to 85C in BIOS and left CTR at 90 (default). I ran diagnostics again and it crashed around 1min after this situation, as you can see with a package temp of 53! (red number in taskbar):
View attachment 2523656

So this is 100% not a thermal problem. While CTR tuning the PC reboots with around 72C, so also this a waaay below critical levels.

Also this is my results after the reboot:

View attachment 2523657

Oh and BTW, I can run CineBench R20 without ANY problems at all with AMD settings (max 83C) and CTR settings (max 74C) and it will not reboot or crash.
Click to expand...
Can you try running the Prime95 small FFT torture test. I have had a situation before where Cinebench ran great, and passed OCCT, but Prime95 small FFT triggered an instant reboot.
 
Save
Reply Quote Rep+ Rep+
#745 · (Edited)
Sleepycat said:
Can you try running the Prime95 small FFT torture test. I have had a situation before where Cinebench ran great, and passed OCCT, but Prime95 small FFT triggered an instant reboot.
Click to expand...
So I just did Tests with lates Prime95. Deactivated all CTR profiles and exited CTR and ran a small FFT Test. This is a screenshot after over 1hr of testing.
2523844



When I start CTR, activate P1 and P2 in hybdrid AVX mode, I get an INSTANT reboot within a second after starting the small FFT test!
2523845


So there must be something ctr related that leads to the reboots.
 
#746 ·
St54Kevin said:
So I just did Tests with lates Prime95. Deactivated all CTR profiles and exited CTR and ran a small FFT Test. This is a screenshot after over 1hr of testing.
View attachment 2523844


When I start CTR, activate P1 and P2 in hybdrid AVX mode, I get an INSTANT reboot within a second after starting the small FFT test!
View attachment 2523845

So there must be something ctr related that leads to the reboots.
Click to expand...
The PX High and Mid profiles you have set in CTR has too low a voltage for the clock speeds you are asking it to do. Prime95 engages these profiles prior to the actual torture test, I suspect for setting up the calculations and expected answers for the test.

Change the two following profiles and run small FFT again.

4.800 @ 1.400 Px Mid
4.900 @ 1.425 Px High
 
Save
Reply Quote Rep+ Rep+
#747 ·
Sleepycat said:
The PX High and Mid profiles you have set in CTR has too low a voltage for the clock speeds you are asking it to do. Prime95 engages these profiles prior to the actual torture test, I suspect for setting up the calculations and expected answers for the test.

Change the two following profiles and run small FFT again.

4.800 @ 1.400 Px Mid
4.900 @ 1.425 Px High
Click to expand...
You didn't look at my Screenshot well enough. PX profiles are not activated because it gives me random reboots in everyday use
 
#749 · (Edited)
@Sleepycat What you suggested let me think about what if CTR has some faulty configs from my broken 5800X. So I delete the config and let it run the diagnostic again. But same picture. This time it crashed again at "manual tuning process" #6 with 1072mV.

Also I just tried to set your suggested values for PX but of course this didnt make a difference as PX profiles won´t be used at more than ~50% CPU load. In that case P1 and P2 gets active and with smallFFT which is 100% load. I really wonder where to search next...
 
#750 ·
St54Kevin said:
You didn't look at my Screenshot well enough. PX profiles are not activated because it gives me random reboots in everyday use
Click to expand...
Ah, yes, I missed that PX was deactivated. My bad.
The only other thing I can think of at this point is an issue with the CTR installation. The Profiles Statistic section of the Profiles page lists every profile's time as 0. That's odd as at least 1 of the profiles in the stats should have a positive number.

I can only suggest to delete your entire CTR installation directory, and unzip a new copy in its place to see if the statistics section starts working again.

Since the stats section says 0, then I wonder if CTR is having trouble determining which profile to use or if it can even detect the load the PC is under, leading to the crash when an extreme load change occurs like with Prime95. FFT small.
 
Save
Reply Quote Rep+ Rep+
#751 ·
Sleepycat said:
Ah, yes, I missed that PX was deactivated. My bad.
The only other thing I can think of at this point is an issue with the CTR installation. The Profiles Statistic section of the Profiles page lists every profile's time as 0. That's odd as at least 1 of the profiles in the stats should have a positive number.

I can only suggest to delete your entire CTR installation directory, and unzip a new copy in its place to see if the statistics section starts working again.

Since the stats section says 0, then I wonder if CTR is having trouble determining which profile to use or if it can even detect the load the PC is under, leading to the crash when an extreme load change occurs like with Prime95. FFT small.
Click to expand...
But It also reboots while undervolting with very little load in diagnostics :D
Oh I played around a bit with the P2 profile and reduced the frequency by 75Mhz. And voila, small FFT test running without problems. I wonder why ctr determined an unstable frequency.
Diagnostic still crashing of course
 
#752 · (Edited)
Sleepycat said:
Ah, yes, I missed that PX was deactivated. My bad.
The only other thing I can think of at this point is an issue with the CTR installation. The Profiles Statistic section of the Profiles page lists every profile's time as 0. That's odd as at least 1 of the profiles in the stats should have a positive number.

I can only suggest to delete your entire CTR installation directory, and unzip a new copy in its place to see if the statistics section starts working again.

Since the stats section says 0, then I wonder if CTR is having trouble determining which profile to use or if it can even detect the load the PC is under, leading to the crash when an extreme load change occurs like with Prime95. FFT small.
Click to expand...
Damn what a stupid solution :D Looks like there was some waste configs in the tool although I clicked the clear config button. Fresh "install" of the tool solved it! No more reboots, both in diagnostic and tuning. Prime95 SMALL FFT still has reboots. I had to lower the P1 profile frequency to get it running stable.
Big thanks! I can now tune further for performance
 
#753 · (Edited)
COD Warzone crashes immediately, rebooting pc when using CTR v2.1RC5 if profiles are autoloaded on boot.
If I disable profiles, reboot pc, start COD Warzone and then enable profiles, gaming is still running and PC does not reboot.

'CHECK STABILITY' instantly reboots PC.

LLC is at 4 all other relevant settings in BIOS at AUTO.
DRAM is running XMP profile.

What causes this?

Font Screenshot Computer Software Electronic device

Font Rectangle Screenshot Technology Software


I also noticed som crazy temperature readings and high volts after I activated CTR profiles and then started playing COD Warzone. Are those just wrong readings from HWInfo or something I should worry about?
Font Rectangle Screenshot Parallel Number
 
#754 ·
FlyinJake said:
COD Warzone crashes immediately, rebooting pc when using CTR v2.1RC5 if profiles are autoloaded on boot.
If I disable profiles, reboot pc, start COD Warzone and then enable profiles, gaming is still running and PC does not reboot.

'CHECK STABILITY' instantly reboots PC.

LLC is at 4 all other relevant settings in BIOS at AUTO.
DRAM is running XMP profile.

What causes this?

View attachment 2524973
View attachment 2524974

I also noticed som crazy temperature readings and high volts after I activated CTR profiles and then started playing COD Warzone. Are those just wrong readings from HWInfo or something I should worry about?
View attachment 2524977
Click to expand...
Your voltages are too low for the clocks that you are asking the CPU to run at. 4700 MHz needs about 1.325 to 1.375V. 4.550 needs 1.200 to 1.225V. Also, you should change the CTR settings mode to advanced, then you can see and set the power and thermal limits too.

As for HWInfo readings, looks weird. The max CPU die average temperature can't be higher than both the max CPU CCD1 and CCD2 die temperatures.

What motherboard are you using?
 
Save
Reply Quote Rep+ Rep+
#755 ·
Sleepycat said:
Your voltages are too low for the clocks that you are asking the CPU to run at. 4700 MHz needs about 1.325 to 1.375V. 4.550 needs 1.200 to 1.225V. Also, you should change the CTR settings mode to advanced, then you can see and set the power and thermal limits too.

As for HWInfo readings, looks weird. The max CPU die average temperature can't be higher than both the max CPU CCD1 and CCD2 die temperatures.

What motherboard are you using?
Click to expand...
The voltages are from running the Diagnostics, so how can they be too low?
Additionally I'm wondering if CPU Tctl/Tdie temps are good. See the new screenshot where it's between 40-76 degrees celcius. I guess that's totally fine.

Why do my HWInfo Current Core Clocks not go any lower than 3700.2 MHz when PX, P1 and P2 Profile are active?

The motherboard I am using is an MSI MAG X570 Tomahawk Wifi with the AGESA 1.2.0.3.b.

Rectangle Font Screenshot Software Electronic device
 
#756 ·
FlyinJake said:
The voltages are from running the Diagnostics, so how can they be too low?
Additionally I'm wondering if CPU Tctl/Tdie temps are good. See the new screenshot where it's between 40-76 degrees celcius. I guess that's totally fine.

Why do my HWInfo Current Core Clocks not go any lower than 3700.2 MHz when PX, P1 and P2 Profile are active?

The motherboard I am using is an MSI MAG X570 Tomahawk Wifi with the AGESA 1.2.0.3.b.

View attachment 2525011
Click to expand...
Something is really not right if those are the settings it chose and you are having stability issues. I had the opposite, I found the settings from diagnostics too conservative and pushed it an additional 100 MHz higher for the same voltage.

Have you uninstalled CTR, deleted the install folder and reinstalled it again? It should go into idle a lot more if you don't have anything running in the background and definitely lower than 3700.
 
Save
Reply Quote Rep+ Rep+
#757 · (Edited)
Sleepycat said:
Something is really not right if those are the settings it chose and you are having stability issues. I had the opposite, I found the settings from diagnostics too conservative and pushed it an additional 100 MHz higher for the same voltage.

Have you uninstalled CTR, deleted the install folder and reinstalled it again? It should go into idle a lot more if you don't have anything running in the background and definitely lower than 3700.
Click to expand...
Just did some gaming and maybe it's magic, but I just played COD Warfare in 2K and Assetto Corsa Competizione in 4K while streaming to YT for a couple of hours with no reboots and no WHEA errors with the same settings. My PC doesn't reboot and games don't crash when gaming, but not having anything running in the background still keeps the idle current core clocks in HWInfo at no lower than 3700 MHz. I deleted the install folder and reinstalled it again and I also changed the NB/SOC LLC to level 4.

I still can't get the Core Clocks in HWInfo to go lower than 3700 MHz, but temps are good 36.8 - 77.1 degrees celcius.
What are your Core Clocks at when 5900X is idle?

Font Line Screenshot Rectangle Software
 
#759 ·
For the guy with unexplained crashes, the author himself said that CTR is made for "gaming" not for AVX heavy tasks, but more and more AAA games use AVX.... and he himself suggests to reduce the max frequency about 50 Mhz in case of heavy AVX or more advanced Prime benching.
 
Save
Reply Quote Rep+ Rep+
741 - 760 of 847 Posts
About this Discussion
846 Replies
360 Participants
Last post:
DorkyZinger84
Overclock.net
posts
28M
members
562K
Since
2004
The premier forum for overclocking experts and enthusiasts. Discuss hardware optimization, custom builds, benchmarking, cooling solutions, and pushing the boundaries of computing performance. From beginner guides to extreme overclocking, join our technical community to master system tuning.
Full Forum Listing
Explore Our Forums
Intel CPUs Hardware News NVIDIA PC Gaming Water Cooling