Home Creators Posts Import Register Favorites Logout

Content

* "TEMP.ini not found or file corrupted" completely fixed. The temporary values are saved in two files. 

* Improved support for low write speed SDDs or "dirty" OS. Avoids situations where logs or settings were corrupted due to a computer reboot. 

* "LibreHardwareMonitorLib" is removed from CTR. The monitoring response speed has increased up to 3 times. Response time during idle time is about 600µs, during 100% load it is 2000µs. 

* The limit for the maximum voltage is now 1.45V. A warning has been implemented if the user is using voltages over 1.35V. 

* Changed PCI mutex hold value. Improves application stability. * Information about the SMU version will be available in google-statistics.  

* HYBRID OC now cannot be enabled during tuning or diagnostics. 

* Found and fixed bug of AVX instruction execution in Zen3 processors (FFT 4096 changed to 2048). This sometimes caused sharp PPT jump, computer reboot or false error in CHECK STABILITY mode. Perhaps this problem also applies to Zen 2. I'll check it out. 

* Information about "CCX VID injection" is also available in diagnostic mode. 

* "CCX VID injection" on some processors with 1 CCX was 0. Partially fixed. However, there are many situations where the motherboard manufacturer has blocked this feature. Hopefully I will get a complete solution to the problem in CTR 2.1 

* Added some tooltips. 

The early access release will take place in a few days. The list will probably be updated if I find bugs or improve anything

Comments

Stefan Tausch

Is it possible to include a button to reset the active OC to stock? i manually start ctr, apply a profile, but when i want to go back to stock behaviour i have to reboot or reset to stock through ryzen master

Riwwelorsch

Hi Yuri, again thanks for the amazing work. Could you please give some info about ctr 2.1? I am interested in hybrid oc 2.0 and auto curve optimizer. Can't wait to try it out

Justin Chao

What is CCX VID injection?

Rares Mihut

Can't wait. RC2 is not great for me.

S. D. Gwynn

For this and previous releases: Is it needed to have Balanced Power Plan or Performance Plan during Diagnoses or will this affect the outcome?

Aragh

I have the same problem, and agree that a reset button would be nice. Currently I start and stop Check Stability to reset the OC within CTR.

SilentGrace

great job!my CHECK STABILITY mode always fall 94.1% with 1.25V 4.6

SANG-HYUN LEE (edited)

Comment edits

2026-09-05 16:00 PMTable read failed! <---what's wrong??
2021-01-21 15:10 PMTable read failed! <---what's wrong??

PMTable read failed! <---what's wrong??

Sakaana303

Should work with any power plan (so it was said at the past).

Rares Mihut

I'm getting Thread# 25 fall down, usage 94.1% on a 5900X 12 core CPU, how can thread #25 be failing if the CPU only has 24 threads?

S. D. Gwynn

Thanks for responding. I think it is necessary to stay on whichever power plan and load line calibration is used for diagnostic and tuning.

S. D. Gwynn

I used Balanced for diag. and tuning also llc5(msi medium). Golden sample 5600x. Hybrid tuning 1 &2. Great results. But when I switched to Performance plan many crashes. Back on Balanced, no problem.

Riwwelorsch

Same here with my 5900X. I thought that it was a simple text error rather than a real thread 25 fell down. But maybe it is being tested with more threads than the processor has ... Yuri?

Riwwelorsch

Does anybody know why CTR is being sometimes loaded with the OS and sometimes not? And I ran the diagnostic three times because of this behavior. At first it was a bronce, than a silver and now Ive got a golden sample. Looks like Zen3 is like wine, gets better with age :-)

1usmus

AMD recommends using the standard power plan (BALANCED).

1usmus

Incompatible BIOS. Sometimes just rebooting the system helps.

1usmus

Read the FAQ please, it says that the category is affected by a huge number of factors. And you probably just have an unstable memory.

Rares Mihut

It happens to my system in diagnostic .

1usmus

I can solve this issue :) thank you for telling me

Rares Mihut

If you need logs I can provide. Thanks for your hard work.

Riwwelorsch

As always when I am dealing with CPU OC I deactivated my memory profile , so it runs with 2133 MHz.

1usmus

I found the reason, but it's only a visual bug. The thread numbers are shifted by 1. So thread 25 is the 24th.

1usmus

It's an interesting situation. In that case, I do not deny the problems associated with AGESA

Riwwelorsch

OK so the visual thing is fixed - Thank you! But why is the test failing at 94,1% with the suggested voltage and frequency? Three people here getting this result. So I think it is unlikly that in all cases the last thread / core failed.

1usmus

This is not a bug. The principle of CTR is to overclock and find an error.

Rares Mihut

The test isn’t failing, it’s detecting performance degradation during the stress test in that particular thread. IE: the performance dropped from 100% to 94.1% , this is how CTR gets stable all core OC.

Riwwelorsch

Thanks for the explanation

Riwwelorsch

Yuri are the default values for P1 (1175 mV) and P2 (1275 mV) your recommendation for longterm usage? Are these voltages more on the safe side? What voltages are you recommend for max performance for a 5900X? At stock the CPU is around 1,30-1,33 during rendering. I also read that MSI issued 1.35 volts (1.30 volts under stress test condition) as a recommendation for manual OC.

1usmus

Hi, my recommendation is a very safe range. The maximum safe voltage is 1.35

Gregoire Pisapia

Is this a maximum safe voltage for Manual OC/Hybrid? AMD is advertising 0,9-1.5V voltage range and basic auto setting (basic PBO) is going freely to 1.5V on boosted cpus in 1 thread application..

Nakano

I'm using a 5950X and I'm unable to use CTR RC2, hopefully this versions works for me, keep the good work!