General Discussions

[hssreddragon]

I'm dealing with some really bizarre reboots on a brand new system I just put together. I've been reading through these discussion forums and seeing a lot of similar issues, some of them years old at this point.

The system in question is:

• AMD Ryzen 7 5800x
• ASUS TUF GAMING X570-PLUS WIFI Motherboard (BIOS v4204)
• G.Skill Trident Z RGB 32GB (2 x 16GB) DDR4-3200 PC4-25600 CL16 Dual Channel Desktop Memory Kit F4-3200C16D-32GTZR
• Corsair RM850X Gold PSU
• Samsung 1TB 970 EVO NVME
• EVGA 1080 FTW3

Settings

• DOCP On
• SVM Mode Enabled
• SR-IOV Support Enabled
• Secure Boot On
• CSM Compatibility Disabled

Tests

• RAM tested in another machine with no problems
• Temps on all hardware are well within specs.
• Video card tested in another machine with no problems.
• Moved NVME to both slot locations to test heat differences
• Installed fresh copy of Windows twice to eliminate possible corruptions from crashes during driver installs.

With a fresh build of Windows 10 21H2 the system will complete the installation process but after running a few Windows Updates will become incredibly unstable. It will start to black screen reboot at random times with no BSOD. It can happen at idle or in the middle of loading drivers and software. The last time I saw something like this is when the B450 motherboards I had were having voltage issues with the AMD 3600 CPUs I was installing. In that case, it was an easy voltage change and move on. This not so much.

Quite frankly this is now like the 4th or 5th major issue I've had since moving to AMD a few years ago and once I complete this build I'm probably going back to Intel. With as many BIOS updates as this board has had in the past year I would expect it to be more stable than this. I have another machine with a 5900x in it with none of these issues. However, I also have a 3600x in another machine that has random lockups with no black screen reboots yet sitting right next to it is a 3600x with no issues at all. All the same motherboards. I build a lot of machines for myself and others and this kind of instability is insane when I'm literally ordering exactly the same supporting hardware. 

I have read and jotted down some things to try based off the comments in this forum. I just wanted to see if there was anyone still having issues currently with this hardware scenario and see if this system is worth trying to save or whether it's just going to be a support nightmare if I turn it over to a new owner. The system currently doesn't stay up long enough to rely on diagnostic software to get a chance to even be installed let along actually run for any amount of time. 

Thank you!

[elstaci]

Possibly you received a defective CPU or Motherboard.

Your RAM seems to be compatible with the 5000 Series processor. I saw the same RAM P/N except instead of a 32GB kit it was a 16GB Kit.  (16GTZR vs 32GTZR)

Try just using 2 sticks of RAM for a Total of 16GBs and see if it continues to shut down. Or just one stick to test only.

Are you able to enter BIOS settings? If so does it show the entire amount of RAM you have installed at the correct RAM Speed?

Have you tried a Clear CMOS to reset BIOS back to factory default settings in case it is a BIOS setting causing your crashes?

I would open a AMD Support - Warranty Ticket and ask them if they believe your processor should be RMAed to be checked by AMD from here: https://www.amd.com/en/support/contact-email-form

I would also open a Asus  Chat session to see if they believe your Motherboard might be at fault and needs to be RMAed.

[hssreddragon]

To address your questions,

1. The RAM kit is only 2 x 16GB sticks so I can remove one and test it but I'm pretty certain the RAM is good since I tested it in another machine with identical motherboard and a 3600x CPU. 

2. I can get into the BIOS without an issue. In fact, the next thing I was going to do is let it sit in the BIOS and see if it reboots or whether it only does it while Windows is up and running. So far I've only seen it reboot when Windows was up but that might just be because I spend so little time in the BIOS beyond the couple settings I mentioned. Windows may also be putting some load on the components that make it fail faster.  Also, I could try to clear the CMOS just to be certain. But again, since I have several machines in the house, I compared most of the settings with them and they are identical. 

3. Yea, I was thinking the same thing on the support. That's mainly why I stopped by here before opening an official support ticket since I wanted to make sure I wasn't missing anything obvious. Thanks for the link! 

Last time I was this frustrated with a motherboard/cpu combo was back in the Slot 1 ASUS/VIA chipset days. I don't feel like reliving it. lol

Thanks!

[elstaci]

The problem is just because one set of RAM is compatible with the 3000 Series processors doesn't mean it will automatically be compatible with the 5000 Series processors.

That is why Asus has a separate QVL List for  RAM for the 3000 and the 5000 series processors.

Also make sure you have the RAM in DIMM slots A2 and  B2.

When in BIOS there is very little stress or activity on any hardware like the CPU or RAM or GPU. It probably wouldn't crash if you left it in BIOS. But when in Windows it starts using parts of the hardware that it wouldn't be using in a BIOS environment. Thus if Windows starts using a part of the Processor or RAM or GPU that is defective it will cause a crash.

But in your case it does sound like a defective processor. Several Users in the past here at AMD Forums received defective 5000 series processors and their various crashes were fixed with a new processor from AMD Warranty.

So I would just see if AMD Support - Warranty believe you may also have received a defective processor.

NOTE: by the way can you keep this thread informed if AMD does say you should RMA your processor and if you receive a new processor and it fixes your crashes will you post here?

[hssreddragon]

Yea, the whole QVL thing was another thing I was looking at as well. ASUS hasn't updated the QVL for this board since like April-June of last year I believe and the model they recommend (GTZA) didn't even show up in a search when I went looking for it. On top of that you have Newegg posting another skew with the last letters of GTZRX as AMD RAM vs the Intel with GTZR.  So, it's more than a little frustrating trying to line up everything perfectly. Honestly, it's been well over a decade since I've had QVL problems with hardware. Generally speaking hardware is very forgiving these days compared to the late 90's early 2000's where RAM is concerned. These new CPUs may change my mind though if this becomes the focal point for the issues. I generally build 4-8 machines every 3-5 years so when I hit a batch of bad hardware it generally happens in bulk. This run has been all over the place.

But yea, I just submitted my service request and once I hear back and with any luck find a solution to this I'll be more than happy to report back on the findings. 

[elstaci]

One thing I learned from being active here at AMD Forums is that Ryzen processors are quite sensitive to the type of RAM it uses.

When I upgraded my AM3+ Asus motherboard to a AM4 Asus TUF Gaming X570 Pro motherboard I did a lot of research on which type of RAM I was going to purchase for my Ryzen 7 3700X processor.

I went to both Asus QVL List for RAM for the 3000 series processor and the RAM's Manufacturer to make sure it also said it was compatible with my Asus motherboard or processor.

Finally I bought 2x16GB G-Skill Trident-Z NEO RAM Set. I far as I am aware I haven't had any issues of crashing or shutting down and I manually XMP the RAM to its native 3600Mhz speed rather than its SPD speed in BIOS.

According to G-Skill RAM CONFIGURATOR your RAM P/N is compatible: https://www.gskill.com/configurator?page=4&cls=1529635169&adSearch2=Memory_Type%C2%A7DDR4%2CCapacity...

So most likely it is a defective CPU or Motherboard.

[hssreddragon]

Well that's good to know! I should have went to the G.Skill site apparently. I forgot all about their configurator page to be perfectly honest. One more thing I can check off the list of concerns. I will test setting things manually though to see how that works out. Had to do that with the i7 950 to get my Dominator RAM to work properly so wouldn't be the first time. The AM4 is my first AMD hardware since the DX2-66 chip so I haven't dealt with their predecessors much where memory is concerned.

Thanks again for the info and the help! 

[elstaci]

You basically have the same Asus type of motherboard except mine is the "Pro" while yours is the "Plus" with the same exact BIOS version I have installed on my,

This was also my first stab at a Ryzen PC built. When I first installed my RAM it was showing at its SPD speed. So I manually set the speed to 3600Mhz and the FCLK at 1800Mhz (half the rate) and everything went smoothly.

Then I decided to use Asus DOCP to make my RAM run at its native speed of 3600Mhz. I noticed there was a DOCP settings in my earlier original BIOS.

That was a big mistake. As soon as I enabled DOCP with 3600Mhz my computer wouldn't boot up and BIOS would reset back to SPD speed.

I then disabled DOCP (Asus  version of  XMP) and again manually overclocked my RAM to its native speed. Has  been working fine and both BIOS and Windows indicates it is running at 3600Mhz.

Then I noticed in the BIOS, In which we both have, I couldn't locate DOCP any more. So it looks like Asus removed setting so that the User can just easily manually set the speed they want from a Drop down menu and the BIOS will automatically do the rest.

EDIT: Forgot to mention do you see any DOCP settings in your BIOS? 

Maybe I missed it. But I went through every settings and couldn't find anything mentioning DOCP.

[hssreddragon]

For reference sake 1/3

 

[elstaci]

Good your BIOS is showing the correct RAM speed of 3200Mhz.

[hssreddragon]

For reference sake 2/3

 

[hssreddragon]

For reference sake 3/3

 

[hssreddragon]

Sorry. Had to split it up due to size and post time limits. 

That's how it's configured now. I just started setting things to manual and I'm running it through a few reboots now to see how it does.

On a sidenote, I purchased all my RAM for all my AMD machines based off the reviews and recommendations that AMD CPU's generally like 3200 speeds. I do know a friend that purchased 3600 Trident Z for his and had such an issue with it that he just bought some 3200 and stopped tinkering at all. For the most part out of the 8-10 machines I've built since I started, 3 of them have had some unusual issues generally involved with voltages or settings that would normally be set to "Auto".  The other 5 machines have been rock solid, have been upgraded once or twice without any issues at all and are all the same hardware except for the CPU's (5900x, 3600x, 3600x, 3600xt). But 3 out of 10 is still a 30% uncertainty when building for others who don't tinker like myself. 

[elstaci]

From researching and answering other Users questions here at AMD Forums I found out that the "Sweet Spot" for RAM Speed for the 2000 Series Processors is 3200Mhz while the "Sweet Spot" for the 3000 Series Processors is 3600Mhz.

I also read the same 3600Mhz Sweet Spot is also valid for the 5000 series Processors. But you can validate

Sweet Spot is when the processor performance is optimized by RAM Speed.

Found this Tech  Review that verifies what I just mentioned above except the Sweet Spot for the 5000 Series is between 3600Mhs and 4000Mhz: https://www.pcguide.com/ram/guide/best-for-ryzen-5000-apu/

For what RAM speeds to choose when picking up some new sticks, we’d recommend doing a little research as to what is the sweet spot when pairing it with your CPU or APU. In the AMD Ryzen 5000’s case, between 3600MHz-4000MHz is that sweet spot and you’ll certainly see some gaming performance gains if you insert RAM of this ilk.

Here is AMD's own RAM Compatibility Chart for their Ryzen Processors: https://www.amd.com/en/processors/ryzen-compatible-memory

[hssreddragon]

Results from last night's experiments.

1. Turned off DOCP and entered settings in manually. - No Change

2. Bumped the voltages on the RAM up incrementally from 1.35-1.395 - No Change

3. Turned the speed of the RAM down to 3000 from 3200.  - No Change

4. Turned off all OC and ran at stock 2133 speeds. - No Change.

Other than the fact I was able to keep the OS alive long enough to pull off a couple Cinebench runs nothing I did seemed to have any effect on the overall stability. It still seems random since it reboots regardless of load. I have a few more things to look at tonight since I was too tired to do any more last night plus I need to respond to tech support with their dxdiag requests. I still have a plethora of other things to try that I've seen in other discussion threads too but I wanted to clear the low hanging fruit tests last night. More to come!

[elstaci]

In the past many Users were having their PC shutdown during idle or any types of load. It was due to a PSU Setting in BIOS.

See in BIOS the following setting is set: Power Supply Idle Control.

From a previous  Reddit thread concerning this setting:

Power Supply Idle Control is for older (or cheaper) power supplies that shut off if the idle wattage is too low. If there's no problems with the PSU only at idle, leave it alone, it'll do nothing for you. Once there's a load on the PSU, it does nothing, and on Ryzen, true idle states are not very common as any program running in the background that monitors the CPU can keep it from idling.

But personally I doubt this is the case but it is a very minor BIOS setting to change to troubleshoot.

[hssreddragon]

Apparently AMD doesn't want me to respond to this thread any further.....

 

[elstaci]

Sounds like a "False"  Spam filtering on AMD software part.

[hssreddragon]

Yea it's strange. I get a "success" posting it followed by spam warning that it's been removed. Tried it a few times even in different browsers with no luck.

[elstaci]

Your last two replies are being posted now.

[hssreddragon]

Great! I'd hate to lose that message since it brings up other questions. Something tells me, there's a duplicate in there if there are multiple responses.

[elstaci]

Most likely a AMD Moderator read your post and adjusted the SPAM Filter to let you post again. just my opinion.

[hssreddragon]

Well, I did follow the link in the spam report to report the error to see if I could draw their attention to it. Kind of odd though that to report an error I have to report it as abuse to fill out the reason but I guess it works as long as they review it. lol

In the end I thought maybe I tripped some hidden maximum character limit with my dissertation on yesterdays events. lol

[hssreddragon]

Update.

Despite constant crashing I was able to upload the dxdiag and the msinfo files that support wanted last night. I received their email overnight with the following suggestions:

"In order to narrow down the issue, I would suggest you to please update system windows 10 to latest build 19044 21H2 since your windows 10 build is 19403 and you may refer the following link to update the windows to latest build:"

After looking at that for a second, I realized they were correct. For whatever reasons Windows has been reporting fully up to date but the build number was not the latest. My mistake. I'll own that since Windows has done that to me before in the past and it seems more interested in pushing Windows 11 these days. So, tried to run the update assistant and of course that was a train wreck because it could never complete. Downloaded the ISO and I'm currently in the process of reinstalling the OS for a 4th time.

Moving on.

"Also I would suggest you to please make sure that you are running your system at stock setting and please disable XMP, PBO(Precision boost overdrive) if it is enable in bios."

Been there, Done that. But fine, they are disabled again.

Now for the one that stirred up so much more fun.

"Once the windows is updated to latest build then please try to update the chipset driver to latest driver version 4.03.03.431 and please visit the following link to download the latest chipset driver"

Oh boy was this a fun one. Well, I've been installing this chipset driver from the beginning and it NEVER installs the full chipset package. AMD power profiles are missing and generally speaking 2 or 3 of the 5 packages never successfully install.

So, just for fun I thought, you know I've been seeing something strange with this on my other machines as well. Let me verify on some 21h2 machines. My findings were that every machine was in some sort of disarray where the chipset package was involved. Meaning on both of the 3600x machines with identical hardware they had different things missing. Like one had the AMD Balanced and Performance modes while the other had just the AMD Performance and no AMD Balanced just for starters. Then when I moved to the 5900x I noticed it had NO AMD power plans and was outright failing to install 2 of the 5 packages at all. Then to top that off I went in to Armoury Crate and snapped this screenshot.

It was then that I realized Armoury Crate doesn't even appear to be installing the correct chipset drivers either suggesting version 3.10.22.706 rather than the latest 4.03.03.431. So naturally, I tried to uninstall and reinstall. Yea, not as easy as it sounds. Even downloaded the AMD software cleanup tool in hopes that it could help and nope. Every attempt to reinstall the current version of chipset would result in 2 of the packages (I believe it was the Ryzen Power Plan and the PCI Device Driver) failing. So now I'm questioning the reliability of Armoury Crate. Again, seen many topics on chipset drivers in these forums that appear to go back 2 years. So my question is this. Should I just ditch AC all together to avoid this in the future? Between Microsoft trying to push drivers and now this I'm feeling that automatic is not the route to go with this platform.

I'll be back after I get done my next reinstall and round of emails.....

[elstaci]

Asus Armoury Crate is installing Asus AMD CHIPSET which may not be the latest version.

Download the latest AMD CHIPSET version for your motherboard from AMD Download page.

Then use MS Install/Uninstall APP to remove all previous Non-Graphics drivers and AMD CHIPSET package from your PC: Microsoft Install/Uninstall Tool 

Then try installing the latest version of AMD CHIPSET and see if it installs correctly with all drivers showing being updated.

I have never had any issues on my Asus TUF Gaming X570 Pro (WiFi) PC installing AMD CHIPSET driver packages in the past and recently.

Except last time I unchecked the Ryzen Power Plan not to install since I rather use Windows Power Plan.

EDIT: Instead of doing a complete clean Windows OS you could have done a In-Place Windows Repair or Update by just going to Microsoft Windows 10 download page and running the latest Windows version on your desktop: Microsoft Windows 10 Download 

That way all your 3rd party Apps and configuration will be automatically transferred to the new Windows folder or your Windows OS will be updated to the latest version without having to install all of your 3rd party apps again.

I personally use this method in the past to fix my Windows OS instead of doing a clean Windows Install. That I will do if the In-Place Repair method doesn't work.

[goodplay]

[hssreddragon]

So Matt, can you confirm which chipset driver is supported? AMD tech support wants 4.03.03.431 and ASUS wants 3.10.22.706 which Armoury Crate and others here seem to agree with at this point. 

The power plan was the visual difference I noticed that something weird was different but I don't feel it's what's causing the main problems with this new build. But it did lead to me to the chipset variations that we're discussing now so it is what it is at this point. 

[elstaci]

Thanks for the update. Next time I go into BIOS I am going to see if DOCP is also listed under AI OVERCLOCKING TUNER.

Strange but I do remember in the latest version which is the same as your BIOS, DOCP wasn't listed. Instead it just had a drop down menu of different speeds to enable.

In the older original BIOS it did show DOCP like the way it was listed in yours.

But I won't use since it didn't work for my 3700X processor and RAM installed.

[filinux]

Hello,

Sorry to hear you are having problems with your build.  I have just quickly gone through the posts here.  I didn't notice any information on the cooling solution you are using for your 5800X or the type of temps you are experiencing during shut down / lock up events.   A lot of what you are describing could be possible due to heat issues.  The Asus TuF X570 (both plus & pro) are GREAT Mobo's.  Not only have they reviewed & tested well online I have had great success with them.  It is actually the MoBo I chose for my son's "Welcome home from the U.S.M.C" complete rebuild that was VERY important to me.  

In troubleshooting I often work with a "what its not" method.  One of the most helpful things I have found over the years is a Linux live distro on a USB stick.  It boots from USB leaving the rest of your system untouched but usually allows you to determine "Is this a Windows issue or a Hardware issue?".   

Hopefully a simple solution will become available to get your new machine up and running at its best.

[hssreddragon]

I use the Scythe Mugen 5 which is on my other machines as well.

https://www.amazon.com/dp/B06ZYB8K77?psc=1&ref=ppx_yo2ov_dt_b_product_details

Have definitely not seen any temps that are even close to what I would consider alarming. Even when I do get a chance to run a Cinebench on it, the highest I've seen is 80c degrees at 100% load. At idle I'll have to check again to be certain but I'd say low 40's generally. Certainly not enough to cause a blackout reboot IMO. Regardless, I'll grab some numbers tonight and verify.

And yea, the ASUS x570 TUF is in my main machine and 4 other systems just in my house, not to mention the ones I've built for others. That's why this situation caught me totally off guard.

[filinux]

Post coffee I noticed that you had included (And I had missed   "All temps are within spec".  

[elstaci]

It is possible that Asus CHIPSET includes not only AMD CHIPSET drivers but maybe some customization for Asus Motherboards which might explain the different versions between Asus and AMD.

The only AMD CHIPSET version that comes close to Asus version is 3.10.08.506 from 10/21/2021 while the version you mentioned is Asus version 3.10.22.706

[hssreddragon]

It's all good Filinux! Best to keep everything on the table. I'd rather double check something than totally miss something. 

After this many years of building you can either be blind to stuff or absolutely paranoid of everything. I think I'm somewhere in between at this point....

 

 

 

[elstaci]

@hssreddragon 

You mentioned you have 4 PCs in your home. I presume you are using it for a Business or do you have a separate PC for each family member?

[hssreddragon]

I have one for each member of my family. I have ones dedicated to office work and video rendering. I have ones for just for gaming. In 1996 we started hosting LAN parties out of my house. One of the primary lessons learned is having people bring their own PC's ended up being hours of troubleshooting vs hours of relaxing and playing games. So in the late 90's early 2000's I started building 8 machines at a time every so many years so that I maintained all of them, kept them on the same build of Windows, kept the games up to date and all that fun. I have enterprise servers in my house as well that help me support this and I use them to demo stuff for my day job. My wife has a laptop for her business but from time to time she'll come down to my office to use the workstations. Tech is my addiction and I'm lucky enough to have a wife that tolerates it. 

[filinux]

Looks like we have a LOT in common!!   

Had to give this post a thumbs up (for a few reasons) and comment because it sparked MANY happy memories of computing past.  The lovely centerpiece in your CPU display, The Pentium Pro, was one of my favorite CPU"s of all time.  Yes, I know this is a AMD Forum, and they have SEVERAL CPU's on that favorite list as well, but I got to keep it real.     I still have my Pentium Pro and 1st Gen Pentium 60Mhz as well tucked away with others.

Showing a much younger techie / gamer friend of mine some old CPU's.  He asked me why Intel didn't just "shove" the Intel 387DX "into the chip".  While explaining it,  I kept thinking "this kid would have a complete breakdown if he had to configure a dipswitch modem in Windows 3.0". 

OH NO!!  I think I am about one sentence away from a full on "Back in my day we had to walk four miles barefoot in a snowstorm to hook up to my local BBS and we were DANG HAPPY to have it!!" 

[hssreddragon]

It's funny you mention the BBS days because my "handle" here is the name I was using in 1986 as a sysop of a local BBS in my area. I used it for the 7 years I was running one and I altered it in 1997 to reflect the gaming group we formed to start playing more "competitively". But yea....like you said, I've survived the era of jumpers, the original AMD vs Intel vs Cyrix wars of the 486 generation, dial-up, IE vs Netscape, Autoexec.bat and HiMem settings....you name it. The CPU display is a badge of honor to me at this point because I used Commodores/Ti-99s through the 80's but it was the x86 platform that I sharpened my teeth on actually building them from parts. It's also what enabled me to work on them the rest of my life rather than just program from magazine articles, tinker with sector editors and game...

Updated the picture and included a fun one since you liked the Pro. When I saw Threadripper I couldn't help myself but take this picture. Took me back to the late 90's again seeing that monster. 

 

Update: Just for clarification too, don't be too hard on me for all the Intel stuff. I just swapped over to AMD in 2019 and all of my Ryzen hardware is still in use so these are only the CPU's I kept that are decommissioned CPU's not active ones. They will make the case when their day comes too eventually.

Oh and for those wondering if there have been any updates on the status of this testing. AMD support basically started the RMA process with last nights test results. But now that the CPU is being taken care of, I'm going to test more on the board and PSU tomorrow to see if there are any more factors that were contributing to the overall problems. If I find anything I will continue to update this thread regardless so it's not left open-ended.

[hssreddragon · Accepted solution]

After a weekend of testing and tearing apart other machines to have enough spare parts to do a full parts swap on this thing I finally isolated the problem down to the CPU. Swapped out PSU, Motherboard and RAM this weekend to finally feel secure in making that call. Once I saw the CPU do the exact same rebooting in a system that had all different components I was done. After all is said and done, the RMA has been submitted, the 3 other machines I have torn apart are back together and I await the processing of the new CPU. This is the first CPU failure, out of the box, that I've seen in a long, long time. I was not expecting it and came here wanting to validate my own knowledge with those in these discussion boards to make sure I wasn't missing something obvious. I appreciate the patience as I thought out loud through the process.

Thanks!

[filinux]

Yep, just like I said, I knew it was a bad CPU all along!!   

O.K., maybe not ALL along, but what is important is you definitely leveled up your already great tech skills solving this one!!

Going on what you said a new CPU being bad out of the box is actually a pretty rare thing in my experience and it is usually one of the last things to actually test (especially if you don't have dedicated test benches set up). 

I have to applaud you further for taking the time to dig in and make sure it was a bad CPU despite it being a  major effort.  I have seen so many people have something simple like a incorrect BIOS setting just going nuts wanting instant RMA and replacement parts.  All those incorrect RMAs (Incorrect installation of CPU's probably costs MASSIVE money in unjustified RMA's) clog up the system and make it harder for people with legitimate faulty hardware and cost us all more in the end.  

I have found AMD to be a responsive company on the very few issues I have had over the years.  Hopefully they get you sorted and happily computing ASAP.  

[hssreddragon]

Yea, like I said, the last time I had a bad CPU out of the box was like the mid to late 90's in my early days of building when hardware was way less forgiving. In the past 3 years I've had 1 bad motherboard out of the box. I just don't see bad hardware that often since I normally stick with the brands that work for me. ASUS, Intel, AMD, EVGA, Corsair. While I don't mind troubleshooting I also don't want to waste a lot of time on hardware that inherently seems broken out of the box. A lesson that young IT often overlook is that time IS money and in my realm MORE valuable than cash. Cash means nothing if you don't have the time to build or troubleshoot what you're working on. 

As for digging in, I'm not one that likes mysteries. I don't like not knowing why something happened or how something works. I've never subscribed to "magic" in my field when dealing with sales reps. As a child I tore apart the most random electronics and spent many hours on Radio Shack electronics kits just to understand how things function. In my adult years I build PC's and fix 80's arcade PCB's. I still enjoy learning things and bringing machines to life or back to life. With this I was not going to be satisfied with just a, "well I got my RMA, so why bother taking it any further?" mentality. It had to be a "hey, what if I still have a bad component in the mix even if the new CPU arrives?". It's always about the next step with me. And yea, I don't want to waste manufacturers time and money any more than my own on returns that aren't necessary which is why I started with the forum here first to check myself before jumping to hardware conclusions. I do think after this experience though that I'm now dedicating a full test bench setup to prevent ripping apart working machines to take a chance with potentially bad hardware. It's something I've looked at for years but didn't want to take the bench space away from my other equipment.