Weird memory timings with Ryzen 5 5600X + freezing issues?

[GreiverBlade]

I wish you could try ECC ram. It would be interesting to see if it reports corrected errors. Perhaps another component is destabilizing your ram?

the issues are the same as i experienced and the initial RAM was the exact same RAM i run

i think spending more on RAM, even more on ECC kits would be wasteful...

but given the OP syspecs with only 2 SSD ... if it's not the OS drive, it could be the MX500?

 

[An0maly_76]

The Patriot kit is listed. But his system is most probably compromised from the frequent crashes caused by the Corsair kit. A Windows reinstall is a must, I say.

Ah, I see now... Methinks I had a typo when I tried it the first time.

One other thing that could be relevant here.

I feel the OP's pain here. My 5900X (all-new build with all-new components) seemed fine for a few months. Then one night the system shut down hard mid-game, fans wide-open. After that, I had no video every 8-12 cold boots, and it didn't seem to wake from sleep either. RGB was going nuts on shutdown too. Troubleshooting was getting nowhere, so Asus suggested a board issue. While it was out for RMA, I bought a cheap B450 board and 3600X to test the other components. The Windows hourglass cursor was a near constant, which hadn't been an issue with the other board.

At someone else's suggestion, I tried my Ballistix DIMMs in solo configuration. BOOM. The second one produced the no-video start. This particular Ballistix SKU was not listed for either board. So I found a Ballistix kit for the AsRock B450 board with a 3000-series, and voila. Hourglass gone, no more no-video startups. Replacing the original kit with a Patriot kit (PVB432G320C6K) from the Asus board's QVL for a 5000-seires cured the 5900X's no-video startup. Still doesn't like sleep mode, but most say sleep mode isn't good for the components anyway.

What's interesting is this. Asus' QVL listed this kit at 16-18-18-36 timings, but DOCP reported some wonky 20-20-20-40 timings. The DIMMs support 3200, but native frequency is 2400. For some odd reason, I had to set manual timings of 16-18-18-36. However, I've since discovered that they run just fine at 2400 with the 20-20-20-40. So this is something else that might be explored here.

Incidentally, my 1700 had some strange issues after using the same Ballistix kit I originally used. For some odd reason, no one ever noticed the BIOS reporting them at different speeds. Yup, not on the QVL for its MSI board either. Found some used HyperX from its QVL, and now it runs perfectly. So it DOES make a difference at least certain cases. Hence, my mention of the QVL, as most don't think about it. I hadn't seen where OP had mentioned this.

Also, while the SSD may have seemed fine up until now, that means literally nothing if it was used the same as a hard drive -- writing, deleting, writing some more, just as most people use HDDs. HDDs fail after so many hours of runtime due to their mechanical nature, regardless of how much or how little is written to them. SSDs are their opposite number -- they fail after so many write cycles.

Something some folks don't realize is that while an SSD's solid-state nature can last for many many years, this only applies if it's used like long-term server storage -- write data that won't be changed for a long time, if ever, and leave it alone.

BIOS version could also be an issue. Asus lists ver 0309 for the 5600X. Which brings me to an interesting observation. They list ver 4404 for the 5600X3D. OP, just to clarify, you are positive this is a 5600X and not some other variant like the X3D, correct? If so, I'm wondering if a memory channel might have gotten flaky. Hope not.

 

[A Computer Guy]

the issues are the same as i experienced and the initial RAM was the exact same RAM i run

i think spending more on RAM, even more on ECC kits would be wasteful...

That's probably true.

but given the OP syspecs with only 2 SSD ... if it's not the OS drive, it could be the MX500?

Ok I reread the thread is this accurate?

- speculating the problem may be ram
- speculating the problem may be OS
- OS reinstalled still no joy
- new RAM (but also not QVL), still a problem with freeze <<< we are here
- maybe one of the SSD's is the problem, try disconnecting both and use a temporary disk ?
- could we be having a power delivery issue, try new PSU ?

 

[GreiverBlade]

just one question, why the QVL for RAM is linked to the mobo? the controller is not on the CPU? i find that a bit weird ... that the exact same RAM, the Vengeance Pro RGB 3600 C18 (ryzen specific with A-XMP ) would behave totally fine with a 3600 and a 5800X3D but not with a 5600X
(that RAM is in the QVL for the MSI MAG B550 Tomahawk i have tho )

to me since the symptome are quite alike, if not totally the same, i would directly think SSD although i did not initially and spent a few hours wondering what the fudge was the issue, until i noticed the SSD, when i unplugged it, it was so hot i could barely touch the aluminium heatsink under the housing

PSU, not sure, it's a Seasonic, my last seasonic, which was a lower end model, lasted me 7yrs and was still perfectly usable when i sold it.

 

[freeagent]

Is that a Chieftec PSU I see in your specs?

Seems a little sus.

 

[A Computer Guy]

just one question, why the QVL for RAM is linked to the mobo? the controller is not on the CPU? i find that a bit weird ... that the exact same RAM, the Vengeance Pro RGB 3600 C18 (ryzen specific with A-XMP ) would behave totally fine with a 3600 and a 5800X3D but not with a 5600X
(that RAM is in the QVL for the MSI MAG B550 Tomahawk i have tho )

Asrock for example lists QVL per CPU type for the motherboard. You can't attach the ram directly to the CPU, need the motherboard for that. The motherboard maker is the vendor thus Qualified Vendor List. The QVL should be representative of the CPU (IMC) + motherboard traces to the ram + RAM kit (at a rated speed and voltage for a model number).

Complication arise of course when ram vendors switch chips in the same model ram.
Also a QVL may not keep up to date with newly manufactured ram or may not likely cover a wide variety of available ram.

You can infer some likelihood of compatibility by comparing a non-QVL ram kit characteristics with a known ram kit on the QVL with similar characteristics. (most importantly the IC used)
Characteristics may include:

• Capacity
• Speed
• Timings
• IC's
• PCB type
• Number of ranks
• Number of dimms

The most important aspect of the QVL in my opinion is that hopefully it has coverage of as many different IC's on the market as possible (only those relevant to the platform of course). The better the coverage of IC's in the QVL the less likely you will have a compatibility issue with non-QVL ram. There are ram settings in the UEFI/BIOS that only the vendor can modify that may be tuned differently for compatibility with different IC's. AMD lays the groundwork for the UEFI/BIOS in AGESA however I'm not sure how much of that compatibility work is on AMD's end or the Vendors end or how much they share in fixing compatibility issues.

PSU, not sure, it's a Seasonic, my last seasonic, which was a lower end model, lasted me 7yrs and was still perfectly usable when i sold it.

Any part can be defective so logically swapping out the PSU is a diagnostic step and often a common one since any kind of bad or incompatible power delivery can cause numerous problems.

 

[GreiverBlade]

Is that a Chieftec PSU I see in your specs?

Seems a little sus.

the OP has a seasonic... (i'm not the OP )

and my PSU is out of any concerne the Proton line is perfectly fine and the review i did read about the Proton 1000w was positive (aslo it's a Sirtec/Sirfa platform, usually they are fine even with Chieftec)
it's the second Chieftec, and the first one i had did last even longer than the Seasonic i mention in my previous post

pros
- Good Rail Stability
- Fully Modular Design
- 80 Plus Bronze Certified
- Electrical Protections (AFC/OVP/OPP/OCP/OTP/SIP/SCP/UVP)
- Peak Output (1155W)
- Price (For Some)

cons
- Warranty (2 Years)
- Noise Levels



i would say the same about EVGA but my issues with them was not with a G+

that SSD death induced freeze was about the only issue i had since i bought that PSU and used it with a R5 3600+GTX 1070 and then RX 6700 XT followed by the actual 5800X3D and RX 7900 XTX

Any part can be defective so logically swapping out the PSU is a diagnostic step and often a common one since any kind of bad or incompatible power delivery can cause numerous problems.

yep, usually BSOD error 41 ... which is what i had with some PSU failing (one LC-Power and one EVGA ) but no "just freeze"


the freeze is symptomatic of a dying/dead SSD or RAM but since it happens with the new sticks too... i wonder

also thanks for explaining the QVL "process"

 

[VuurVOS]

That's probably true.

Ok I reread the thread is this accurate?

- speculating the problem may be ram
- speculating the problem may be OS
- OS reinstalled still no joy
- new RAM (but also not QVL), still a problem with freeze <<< we are here
- maybe one of the SSD's is the problem, try disconnecting both and use a temporary disk ?
- could we be having a power delivery issue, try new PSU ?

Its not hardware. This problem is caused by operating system or the software installed. The real question is, which one is causing it.
I already hammering on that for some time with my posts
https://www.techpowerup.com/forums/...n-5-5600x-freezing-issues.315089/post-5130517
https://www.techpowerup.com/forums/...n-5-5600x-freezing-issues.315089/post-5130594
https://www.techpowerup.com/forums/...n-5-5600x-freezing-issues.315089/post-5130623
https://www.techpowerup.com/forums/...n-5-5600x-freezing-issues.315089/post-5130851

Could be a Windows patch, or even Asus software which is installed later on:

New ryzen build keeps freezing and I am at my wits end

I just built a new ryzen system which I have been having trouble with ever since. Essential components are CPU: Ryzen 5600x Mobo: Asus B550F Gaming RAM: 4x 8gb Crucial Ballistix 3600 MHz DDR4 GPU: Nvidia 1060 What happens is that the picture just freezes and I can only use the mouse cursor...

arstechnica.com

@cst1992
My advice:

1. Return the memory you have bought within the cooling-off period.
2. Write down all the software you have currently installed. Also the Windows patches and share it with us.
3. Do a clean install of Windows and DONT INSTALL OTHER SOFTWARE! Keep it clean as long as possible without any OS "tweaks".
4. After sometime without freezes, just install one piece of software you had installed earlier and write down what you have installed.
5. Wait for sometime and then repeat step 4 when you have no freezes.

Off topic:
I spent more than month to find out that you shouldn't rename the hostname of a laptop which is AzureAD joined since it will break the Edge Browser There was no clear eventlog message. Did a couple of reinstalls and found the root cause by doing trial and error.

 

[cst1992]

Thank you everyone for your support.

Which software do you currently have installed on your system? Are Windows updates installed?

Yes, updates began installing as soon as I connected to the Internet after installing the system.

I have the exact same memory kit (well, the 2x 32GB one) on my 5900X + X570S Aorus Pro AX and never had one iota of trouble with it.

I bought it mainly because of the Corsair thing and the fact that it was on the QVL. Also, Tom's did a review of the kit you have and they had good things to say about it.

After installing the new kit?

Progression is as follows:
Install 1: Windows 10 Pro, 2 September 2023, Corsair modules (non-kit, DOCP)
Install 2: Windows 11 Home, 26 October 2023, Corsair modules (DOCP, then base frequency)
Install 3: Windows 11 Home, 21 November 2023, Patriot modules (DOCP).
I did install 3 because I got that issue with the wireless card, so I thought it was best to nuke it all and go with reset CMOS, plus new system and new modules.
Wireless card issue came back after reboot.

the issues are the same as i experienced and the initial RAM was the exact same RAM i run

Yours is the RGB Pro (which I am expecting is a better-binned Corsair kit) and mine was the LPX (which is average).

but given the OP syspecs with only 2 SSD ... if it's not the OS drive, it could be the MX500?

The MX500 is barely used. It was in a drawer for a year (as I bought it in a sale) then when I built the system I installed it as a high-speed storage drive.
Well, to be more accurate it wasn't completely unused, I used it to store some of the excess files from my laptop once or twice in the said year.

I tried my Ballistix DIMMs in solo configuration. BOOM. The second one produced the no-video start.

The individual Corsair modules booted fine, except if used at 3600MHz in the B1 slot.

Replacing the original kit with a Patriot kit (PVB432G320C6K) from the Asus board's QVL for a 5000-series cured the 5900X's no-video startup.

So at least I shouldn't be getting a RAM error when booting with the new kit, even if I tried to use individual modules. Still no guarantees about the B1 slot (why try it if it is not recommended?), but a normal boot should always happen.

What's interesting is this. Asus' QVL listed this kit at 16-18-18-36 timings, but DOCP reported some wonky 20-20-20-40 timings. The DIMMs support 3200, but native frequency is 2400. For some odd reason, I had to set manual timings of 16-18-18-36. However, I've since discovered that they run just fine at 2400 with the 20-20-20-40. So this is something else that might be explored here.

DOCP timings are fine - you can check the ZenTimings screenshots above.

Something some folks don't realize is that while an SSD's solid-state nature can last for many many years, this only applies if it's used like long-term server storage -- write data that won't be changed for a long time, if ever, and leave it alone.

I am not getting any values that concern me in SMART. If any of the two SSDs is bad, how do I tell? Heat is not an issue as both slots for NVMe drives on this motherboard come with a heatsink on them.
I could try creating a partition on the MX500 and installing Windows on that. Tried so much so far, what's one more? But in that case, should I just remove the NVMe drive from the board?

BIOS version could also be an issue. Asus lists ver 0309 for the 5600X. Which brings me to an interesting observation. They list ver 4404 for the 5600X3D. OP, just to clarify, you are positive this is a 5600X and not some other variant like the X3D, correct? If so, I'm wondering if a memory channel might have gotten flaky. Hope not.

I'm pretty sure as I bought a sealed box. As of Install 2 I am using version 4802 of the BIOS, so in theory it should support the 5600X3D as well.

Ok I reread the thread is this accurate?
- speculating the problem may be ram

Yes, now using a new kit.

- speculating the problem may be OS
- OS reinstalled still no joy

Yes, plus reinstall. The speculation happened when Windows 10 was still on there, so the system has been reinstalled twice since then.

- new RAM (but also not QVL), still a problem with freeze <<< we are here

No. New RAM is in the QVL.

- maybe one of the SSD's is the problem, try disconnecting both and use a temporary disk ?

I don't have a spare - though I could disconnect one of them and use the other alternately and see if anything changes.

- could we be having a power delivery issue, try new PSU ?

Any part can be defective so logically swapping out the PSU is a diagnostic step and often a common one since any kind of bad or incompatible power delivery can cause numerous problems.

I've been using the power supply in my old build for years - and would have if I hadn't decided to build this system. Why would it suddenly go bad?
I have seen the voltages in BIOS and HWInfo64. Didn't see anything weird.

just one question, why the QVL for RAM is linked to the mobo? the controller is not on the CPU? i find that a bit weird ... that the exact same RAM, the Vengeance Pro RGB 3600 C18 (ryzen specific with A-XMP ) would behave totally fine with a 3600 and a 5800X3D but not with a 5600X
(that RAM is in the QVL for the MSI MAG B550 Tomahawk i have tho )

to me since the symptome are quite alike, if not totally the same, i would directly think SSD although i did not initially and spent a few hours wondering what the fudge was the issue, until i noticed the SSD, when i unplugged it, it was so hot i could barely touch the aluminium heatsink under the housing

SSD has a heatsink on it, and temperatures are okay in HWInfo - high 40s or low 50s.

PSU, not sure, it's a Seasonic, my last seasonic, which was a lower end model, lasted me 7yrs and was still perfectly usable when i sold it.

Mine is a Gold-rated Seasonic. Not high-end like those Platinum or Titanium units, but it has been rock solid so far.

Could be a Windows patch, or even Asus software which is installed later on:

I have no ASUSware installed in my system. I was aware that Armoury Crate messes with the system's kernel, so I installed all the drivers from manufacturer's websites (for wired/wireless, graphics, hard drives) or AMD's website(chipset).

My advice:
Return the memory you have bought within the cooling-off period.

It's not returnable. https://www.amazon.in/gp/product/B088KSRW4S
In any case I was thinking of keeping this as it's a kit (unlike the Corsairs) and on the QVL.

Write down all the software you have currently installed. Also the Windows patches and share it with us.

If you mean updates, there's just two: KB5032007, KB5032190.
As for software, there's just drivers, utilities eg. Cinebench, CPU-Z, CoreTemp, HWInfo, Afterburner, Git, Cheat Engine, VLC Media Player
Third party software: L-Connect 3 (for fans), Auslogics Boostspeed 13, Quick Heal Total Security (my antivirus).
Other stuff is my Steam games, an old game (NFS Hot Pursuit, the 2010 one)

Do a clean install of Windows and DONT INSTALL OTHER SOFTWARE! Keep it clean as long as possible without any OS "tweaks".
After sometime without freezes, just install one piece of software you had installed earlier and write down what you have installed.
Wait for sometime and then repeat step 4 when you have no freezes.

How do I tell? As we've talked about before, the freezes are arbitrary.

Off topic:
I spent more than month to find out that you shouldn't rename the hostname of a laptop which is AzureAD joined since it will break the Edge Browser There was no clear eventlog message. Did a couple of reinstalls and found the root cause by doing trial and error.

My system is not connected to Azure. I used Rufus to make my bootable media, and I am using a local account. I did rename my system, but only because it had some ugly name like DESKTOP-XXXXXXX.

---

Okay, so what I figure is that there are two issues here, at least.
1. Something is causing the system to freeze, something which is still happening, in spite of RAM change and system reinstalls.
2. I am getting issues with stuttering in my games, none at first, then more and more until the time the slow freezes start to happen and I have to reinstall the system. It's pretty much a given that the slow freezes (meaning the system is responsive at first, but then starts lagging at even a context menu, to the point of eventually becoming completely unresponsive) is the result of a corrupted OS. The new kit may or may not help with the OS corruption, or it may simply be a result of the hard reset I have to do after a freeze, as the system simply doesn't recover from said freeze.

One piece of news I have is that the wireless card error I am having persists after a reboot. A driver reinstall fixes the issue until the next reboot. Now it's even happening that I am getting connectivity errors and network drops. This card worked fine for months - why die so early?

ZenTimings doesn't show a voltage value for the new kit either. This is technically a BIOS issue, but I am on the latest BIOS as of the start of this month (new version is for fixing the Inception vulnerability). Could this instead indicate a motherboard issue?

I have not disabled C-State Control in BIOS (like some people recommend to fix freezing), as I wanted to keep as many settings default as possible. I have, however, disabled motherboard RGB and enabled resizable BAR. I have also not modified any settings in BIOS related to power supply current which some people recommend to change from low power idle to typical.

If the issue is nothing else, could it be the CPU? I ran corecycler as recommended by @VuurVOS but it froze before it could go past the 12 hour mark. It didn't report any issues in said 12 hours.

Also, if anyone has any idea what HDD Power Saving is and how to disable it, let me know.

Power saving features for HDD and CPU seem to be a root of many issues in these new motherboards, it seems.

 

[wNotyarD]

Thank you everyone for your support.

Yes, updates began installing as soon as I connected to the Internet after installing the system.

I bought it mainly because of the Corsair thing and the fact that it was on the QVL. Also, Tom's did a review of the kit you have and they had good things to say about it.

Progression is as follows:
Install 1: Windows 10 Pro, 2 September 2023, Corsair modules (non-kit, DOCP)
Install 2: Windows 11 Home, 26 October 2023, Corsair modules (DOCP, then base frequency)
Install 3: Windows 11 Home, 21 November 2023, Patriot modules (DOCP).
I did install 3 because I got that issue with the wireless card, so I thought it was best to nuke it all and go with reset CMOS, plus new system and new modules.
Wireless card issue came back after reboot.

Yours is the RGB Pro (which I am expecting is a better-binned Corsair kit) and mine was the LPX (which is average).

The MX500 is barely used. It was in a drawer for a year (as I bought it in a sale) then when I built the system I installed it as a high-speed storage drive.
Well, to be more accurate it wasn't completely unused, I used it to store some of the excess files from my laptop once or twice in the said year.

The individual Corsair modules booted fine, except if used at 3600MHz in the B1 slot.

So at least I shouldn't be getting a RAM error when booting with the new kit, even if I tried to use individual modules. Still no guarantees about the B1 slot (why try it if it is not recommended?), but a normal boot should always happen.

DOCP timings are fine - you can check the ZenTimings screenshots above.

I am not getting any values that concern me in SMART. If any of the two SSDs is bad, how do I tell? Heat is not an issue as both slots for NVMe drives on this motherboard come with a heatsink on them.
I could try creating a partition on the MX500 and installing Windows on that. Tried so much so far, what's one more? But in that case, should I just remove the NVMe drive from the board?

I'm pretty sure as I bought a sealed box. As of Install 2 I am using version 4802 of the BIOS, so in theory it should support the 5600X3D as well.

Yes, now using a new kit.

Yes, plus reinstall. The speculation happened when Windows 10 was still on there, so the system has been reinstalled twice since then.

No. New RAM is in the QVL.

I don't have a spare - though I could disconnect one of them and use the other alternately and see if anything changes.


I've been using the power supply in my old build for years - and would have if I hadn't decided to build this system. Why would it suddenly go bad?
I have seen the voltages in BIOS and HWInfo64. Didn't see anything weird.

SSD has a heatsink on it, and temperatures are okay in HWInfo - high 40s or low 50s.

Mine is a Gold-rated Seasonic. Not high-end like those Platinum or Titanium units, but it has been rock solid so far.

I have no ASUSware installed in my system. I was aware that Armoury Crate messes with the system's kernel, so I installed all the drivers from manufacturer's websites (for wired/wireless, graphics, hard drives) or AMD's website(chipset).

It's not returnable. https://www.amazon.in/gp/product/B088KSRW4S
In any case I was thinking of keeping this as it's a kit (unlike the Corsairs) and on the QVL.

If you mean updates, there's just two: KB5032007, KB5032190.
As for software, there's just drivers, utilities eg. Cinebench, CPU-Z, CoreTemp, HWInfo, Afterburner, Git, Cheat Engine, VLC Media Player
Third party software: L-Connect 3 (for fans), Auslogics Boostspeed 13, Quick Heal Total Security (my antivirus).
Other stuff is my Steam games, an old game (NFS Hot Pursuit, the 2010 one)

How do I tell? As we've talked about before, the freezes are arbitrary.

My system is not connected to Azure. I used Rufus to make my bootable media, and I am using a local account. I did rename my system, but only because it had some ugly name like DESKTOP-XXXXXXX.

---

Okay, so what I figure is that there are two issues here, at least.
1. Something is causing the system to freeze, something which is still happening, in spite of RAM change and system reinstalls.
2. I am getting issues with stuttering in my games, none at first, then more and more until the time the slow freezes start to happen and I have to reinstall the system. It's pretty much a given that the slow freezes (meaning the system is responsive at first, but then starts lagging at even a context menu, to the point of eventually becoming completely unresponsive) is the result of a corrupted OS. The new kit may or may not help with the OS corruption, or it may simply be a result of the hard reset I have to do after a freeze, as the system simply doesn't recover from said freeze.

One piece of news I have is that the wireless card error I am having persists after a reboot. A driver reinstall fixes the issue until the next reboot. Now it's even happening that I am getting connectivity errors and network drops. This card worked fine for months - why die so early?

ZenTimings doesn't show a voltage value for the new kit either. This is technically a BIOS issue, but I am on the latest BIOS as of the start of this month (new version is for fixing the Inception vulnerability). Could this instead indicate a motherboard issue?

I have not disabled C-State Control in BIOS (like some people recommend to fix freezing), as I wanted to keep as many settings default as possible. I have, however, disabled motherboard RGB and enabled resizable BAR. I have also not modified any settings in BIOS related to power supply current which some people recommend to change from low power idle to typical.

If the issue is nothing else, could it be the CPU? I ran corecycler as recommended by @VuurVOS but it froze before it could go past the 12 hour mark. It didn't report any issues in said 12 hours.

We've been through so much troubleshooting, and we don't seem to be much closer to a solution. Ouch.
If you're still having issues with a specific memory slot (B1) though, either the motherboard or the CPU (after all, the controller is on it and it has been running memory off-spec for a while) is bogus.

 

[TheoneandonlyMrK]

I get similar if running 1900 fabric and memory for 3800 MHz at 18,22,22,22,42,64.

Works perfectly so it seams then now and again it crashes.

Lowering speed helps , try it , if it works it's your IMC

Then at least you know where to focus your attention.

ATM this thread is all memory.

Yet the IMC is 50% of this equation.

Also memory in A2 B2, there's a good reason.

Using a1or b1 would leave unfilled socket and circuit after the memory which would be recieving signals.
This can cause signalling issues due to ringing etc.

There's a pbo thing that makes Ryzen 5#### miss behave too pbo fmax enhancer turn that off.

Not sure you can pbo at all on x3d chips tbf.

 

[VuurVOS]

It's not returnable. https://www.amazon.in/gp/product/B088KSRW4S
In any case I was thinking of keeping this as it's a kit (unlike the Corsairs) and on the QVL.

Shame it is non-returnable

If you mean updates, there's just two: KB5032007, KB5032190.
As for software, there's just drivers, utilities eg. Cinebench, CPU-Z, CoreTemp, HWInfo, Afterburner, Git, Cheat Engine, VLC Media Player
Third party software: L-Connect 3 (for fans), Auslogics Boostspeed 13, Quick Heal Total Security (my antivirus).
Other stuff is my Steam games, an old game (NFS Hot Pursuit, the 2010 one)

Can you remove Quick Heal Total Security since Windows already have a virus scanner. Please remove Auslogics Boostspeed 13, Cheat Engine and L-Connect 3, Afterburner as well.
Which version of CPU-Z, CoreTemp, HWInfo are you using? Are those portable editions?

Maybe it is wise to do a clean install of Windows with the latest Windows patches & drivers. With only steam and git besides the operating system. Nothing else.

 

[cst1992]

Lowering speed helps , try it , if it works it's your IMC

Haven't tried it on this kit, but for the Corsair modules I got a crash at the base frequency of 2666 MHz as well. Do you think it could be because they were not a kit? If so it surprises me how poorly this memory controller is designed.

There's a pbo thing that makes Ryzen 5#### miss behave too pbo fmax enhancer turn that off.

It's set on Auto at the moment. Is that good enough or should I explicitly turn it off?

Not sure you can pbo at all on x3d chips tbf.

I don't have an X3D chip. They are only in the US and then only at Micro Center. I have a regular 5600X.

Can you remove Quick Heal Total Security since Windows already have a virus scanner. Please remove Auslogics Boostspeed 13, Cheat Engine and L-Connect 3, Afterburner as well.
Which version of CPU-Z, CoreTemp, HWInfo are you using? Are those portable editions?

Boostspeed isn't running all the time. L-Connect is, but I need that to control the fan lighting (it's too bright at 100%) so I cannot permanently remove it.
AFAICT none of them are portable editions - I used installers for all of them.

Maybe it is wise to do a clean install of Windows with the latest Windows patches & drivers. With only steam and git besides the operating system. Nothing else.

Do you think a specific software is the culprit?

 

[TheoneandonlyMrK]

Haven't tried it on this kit, but for the Corsair modules I got a crash at the base frequency of 2666 MHz as well. Do you think it could be because they were not a kit? If so it surprises me how poorly this memory controller is designed.

It's set on Auto at the moment. Is that good enough or should I explicitly turn it off?

I don't have an X3D chip. They are only in the US and then only at Micro Center. I have a regular 5600X.


Boostspeed isn't running all the time. L-Connect is, but I need that to control the fan lighting (it's too bright at 100%) so I cannot permanently remove it.
AFAICT none of them are portable editions - I used installers for all of them.

Do you think a specific software is the culprit?

definitely try and explicitly set it to off, you can use pbo,, just that enhancement is the issue.

as for your first reply,it could have been, that single ram sticks are an opportunity for issues but so is the above, who then would know if it were one or the other causing issues.

i get LESS random crashes with that disabled, none at my stable 3600 settings tbf, I do have custom water though.

jtbc i would try that setting.


after that if issues persist slow ram speeds.

 

[VuurVOS]

Boostspeed isn't running all the time. L-Connect is, but I need that to control the fan lighting (it's too bright at 100%) so I cannot permanently remove it.
AFAICT none of them are portable editions - I used installers for all of them.

You can still use the system without the L-Connect software. Just temporally cover the Window of the case with a blacket so you are not blinded by the fan lighting while troubleshooting or use normal fans without RGB if you have those lying around in a box

Do you think a specific software is the culprit?

Problem is software. I am 99,9% sure of it. The question is which software is causing it. At this moment I really dont know.
Thats why I am hammering on keeping the installed software to a bare minimum without any third party software.

If you install lots of different software again which you really dont need, then it will be a pain in the ass to find out what is causing it.
Keep track on a piece of paper which software you install and when you install.

If it is happening for example like three times a week at this moment, then the first target with a clean install will be to survive a week without any freezes with a bare minimum of software installed.
After that you can start adding one piece of software which you want to have on your system.
Everytime you install something, you need to write it down which software and when it is installed so you have a tracing log. Then you wait again for a week before installing more software.

If it pops up after a month then you know what you have installed something last week and that could be causing the problem.

In short my suggestion for now is:

1. Clean Windows Installation. No third party software.
2. Only install drivers which are provided by the motherboard, graphics card manufacturer. Dont install control software for keyboard, mouse, headset.
3. Install steam with a game you often play
4. Wait until you are certain the freezes doesn't happen, then install one additional software you really want to have on your system
5. Repeating the previous step over and over again.

 

[cst1992]

Before I do another system wipe, I want to try something: installing all drivers from the Asus motherboard page.

So far I have been downloading drivers from manufacturers' sites (e.g. Intel and Realtek) and installing the latest versions of those.

What if I installed an incompatible driver without realizing it?

I'll report back if I see a change.

Edit: I found I had never installed my audio drivers. But yeah, apart from that - no change.

There's a pbo thing that makes Ryzen 5#### miss behave too pbo fmax enhancer turn that off.

Currently testing this.

 

[remixedcat]

@phanbuey I love that term used "POSSESSED"...hehehe a doom machine
Im not an AMD user but, i also agree that i would go Kingston or GSkill if i was setting up an AMD platform

My old phenom system had a ton of issues w corsair xms ram too.. switched to gskill n it fixed it. Kept having server 2008 lock up randomly or hang

Oh .. and STOCK TIMIMGS!!

Glad I saw this thread coz I'm saving up to complete a ryzen build... only got mobo, gpu, and ssd for it so far... I know now to stay away from corsair..

Also they blocked me on some socials when I told them about my ram issue.. like wtf...

 

[agent_x007]

1) I would try forcing [100.000MHz] on BCLK from [Auto] setting (in BIOS/UEFI).
2) Instead of checking for just RAM errors with TM5, check HWInfo64 for WHEA errors (it's counter is at the end of list of parameters it shows)

 

[cst1992]

WHEA counter is at zero.

 

[agent_x007]

OK. I mentioned locking BCLK because of this video BTW :

EDIT : I had "cycling freezes" (mouse/progams freeze for a second or two, and after that it's fine for a minute or two), when CLDO VDDP and VDDG CCD/IOD voltages were set too low during my RAM/IF OC.

 

[cst1992]

I think the DOCP profile sets BCLK to 100.0000 already.

 

[izy]

My RAM seemed perfectly fine till i had some random reboots and poor performance and ive tested it with OCCT RAM stress test and it was starting to spit errors after 20 mins, TM5 was error free (tested overnight), went from CL16 to CL 18 and it fixed the problem so far, i have 2 x 16GB kits in my computer and one is rated cl16 and one cl18 and i was running them tunned at CL16 3733mhz seems that the second kit doesnt like CL16, using only the first kit i could even run CL14 @ 3733mhz gdm off.
You should test it with OCCT RAM TEST, seems that it detects unstable RAM better.

 

[The King]

Try downloading this software. After installing it click the Analyze button. If you see memory management errors then it could be RAM related.
If you see WHEA errors it could be CPU related.

Resplendence Software - WhoCrashed, automatic crash dump analyzer

WhoCrashed, automatic crash dump analyzer for Windows

www.resplendence.com

 

[VuurVOS]

My RAM seemed perfectly fine till i had some random reboots and poor performance and ive tested it with OCCT RAM stress test and it was starting to spit errors after 20 mins, TM5 was error free (tested overnight), went from CL16 to CL 18 and it fixed the problem so far, i have 2 x 16GB kits in my computer and one is rated cl16 and one cl18 and i was running them tunned at CL16 3733mhz seems that the second kit doesnt like CL16, using only the first kit i could even run CL14 @ 3733mhz gdm off.
You should test it with OCCT RAM TEST, seems that it detects unstable RAM better.

Its not RAM! Also mixing obviously two different memory kits (CL16 + CL18) is asking for issues. For a reason the second kit is rated CL18 instead of CL16
@cst1992 Happy Birthday!

 

[cst1992]

Thanks!

No matter what anyone says, I'm not getting any more RAM.

The fmax enhancer thing has reset my CPU's max all-core clock to 4.1 GHz and results in a 10% hit in Cinebench scores. If it fixes this issue though I don't mind in the slightest.

Try downloading this software. After installing it click the Analyze button. If you see memory management errors then it could be RAM related.
If you see WHEA errors it could be CPU related.

What if there are no dumps?

In a sense these are the cleanest crashes I have seen - they leave no droppings.