Mr. Fox

Chances are high that the hardware is OK and the problems revolve around a firmware dumpster fire covered in feces. The best hardware in the world is as worthless as the tits on a boar if the firmware is junk. You are not going to win any races in a Top Fuel dragster with a malfunctioning transmission no matter how much horsepower it has.

I said to ignore part 1. Maybe you should go watch part 1. Luumi does not demonstrate it as GamersNexus did, but he talks about that process sometimes damaging the modules. It might be more fun to do it that way though. The way I have always done it works and causes no damage to the memory but it ruins the original adhesives and pads in the process. If you need to sell or RMA the memory and any point that part is less than ideal because it makes putting them back together more difficult and costs something more than zero.

Setting the water block on top with no thermal pad (direct metal-to-metal contact) dropped the temps yet another 3°~4°C instantly.

Wow, that worked flawlessly. I put my memory sticks in laquer thinner standing on the connector edge in a shallow plastic container so that the specification label (the one that has "warranty void if removed" on it, LOL) was not submerged or damaged by the solvent. I made the thinner deep enough to reach all of the the adhesives though not deep enough to reach that label. Within about 5 minutes I was able to easily separate the heat sink from the modules. I did it gently so that the original adhesive remained on the heat sinks and slowly peeled away from the PCB. After letting them dry, I put waxed paper on the adhesives and pads so that if I ever need to RMA the modules I can reinstall the stock heat sinks and there will be no visible evidence of removal or modification. All of the factory adhesives and thermal pads are exactly where they belong on the heat sinks. Now I have the water cooling jackets back on them again and they are much, much cooler, even without the water block. Temps dropped from ~51°C to ~42°C in the TM5 1usmus_v3 test. This is enough improvement to mean the difference between zero errors and failing the stress test and instability due to high memory temps running identical memory overclock settings.

Easy heat sink removal... ignore part 1 video. Such a funny maniacal laugh. I have been doing it the hard way. This should also require less cleanup effort.

I love watching this young lady pound away on those drums. She is really good at it.

Here's a new one from Brother Tom... @ryan Nice to see something new from Disturbed, too.

Looks like the internet gremilns found your router. 😉

Yup, I actually already did that. The only reason I waste any drive space on any version of Windows now is because I enjoy benching and using whatever OS works best. Apart from that hobby, Linux meets nearly all of my OS needs as well as Windows does, and better in some ways. For work, the only reason I need to use Windows is because I have to use PowerBI, and the alternatives to Micro$lop Office universally suck, especially Outlook. It has no acceptable substitute. Otherwise, LibreOffice Write and Calc are essentially adequate as replacements for Word and Excel. If I could bridge that gap, I would probably even seriously consider walking away from the benching hobby and find another form of entertainment.

I'm going to have to take the position that if 4090 and beyond won't have working Windows 7 drivers I will not be looking at a GPU upgrade for a long time. At least not for my benching rig. I might get a 3090 to upgrade from my 2080 Ti someday. Maybe when I can get a nice used 3090 for $500-750. The notion of giving up Windows 7 permanently doesn't register anywhere on my radar at this time.

Just to be clear, he is using a new version of TB that is not 100% working with DDR5 and the versions we have access to does not. I hope he has a version of TB for DDR5 that we will be able to get soon so we don't have to worry about this. Save a copy of the memory firmware and we're good for recovery if this is a common DDR5 issue once we have a DDR5-compatible version of TB. I hope yours is not messed up, too. Given that one stick of the Corsair stopped booting, I am tempted to reset BIOS defaults and drop back down to A30 or A40. MSI can pretend the problem is somehow a failure on TeamGroup's product, but there is no way it was a coincidence. It might have been a one-off glitch that is rare, but the memory didn't decide to corrupt itself at exactly the moment I flashed A60, LOL.

I already did that, but I did not copy/paste my reply to MSI when I posted their response. I gave them a copy of the invoice for replacement memory and offered to send them the corrupted modules and their response was to RMA the memory, LOL. They have not responded (yet) to the information I shared with them for Vitaliy's email. I will not be surprised if they do not respond.

OK. As expected, MSI passed the buck and shirks their responsibility. Typical loser OEM, just like ASUS. "Generally speaking, BIOS update does not cause damage to memory SPD..." But, let's have TeamGroup pay for it... nice solution for MSI. EVGA FTW! Our friend at Thaiphoon Burner replied. He is such a great guy... and, very smart. ------------------------------------ From: Vitaliy Jungle Sent: Tuesday, July 19, 2022 2:57 AM To: Mr. Fox Subject: Re: "Thaiphoon Burner PRO Corporative Edition" Hello Mr. Fox, Thank you for the reports sent! I have examined the SPD dumps from your Team Group memory modules. They are both corrupted indeed. However the first module with the corrupted Part Number "ŠEAMGROUP-UD5-620›" has only 1 error in the first 512 bytes. This means DRAM and module organization, timings, voltages and other electrical parameters are not corrupted. The Part Number also needs to be fixed. As to the second SPD dump, it has more random errors in the first 512 SPD bytes. Both of the SPD dumps have a corrupted XMP header. This is why BIOS doesn't find XMP profiles. Here is how the XMP header looks like in the SPD of your Corsair modules: It is not corrupted like the other part of XMP bytes. And here is the XMP header of your Team Group modules: The header is corrupted. As you can see many of the XMP bytes are set to "20". This is a "blank" code and it is related to a string code. And it is not an error. So, if the Thaiphoon Burner was ready, you could restore your SPD easily. But at this moment I am unable to help you. Anyway, thank you for sending me reports! I felt much better when I saw that my program is already capable of decoding some parts of DDR5 SPD without issues. With best regards, Vitaliy. @johnksss@Rage Set

Yes, I did it as Brother @jaybee83 suggested. Packed it in from the large open end.

I am sorry to hear that. I am not sure what to suggest at this point.

Got an new ssd to put in more OS's. From what I have read. After 5= quintuple boot then we will have 6 = sextuple, 7 = septuple, 8 = octuple 🙂 You probably already know this but for the benefit of anyone else reading that may not, it is always best to install the OS with only the drive you wish to install it on. All others should be physically removed or disabled in the bios. That way each operating system will have its own unique BCD/MBR. Then you can add and remove anytime you want to without having to worry about whether or not it will affect one of the other installed operating systems. If you install multiple Windows version with a previous there will be only one bootloader on one drive, and the function of all OSes will be dependent upon the presence of the drive where the first OS was installed because the bootloader for all of them is on that drive. @Tenoroon if this is not the one I shared with you before, try it first. XTU 4.2.0.8 AW18 This is probably the one I gave you before, but I do not recall for certain.

It was set to anyone with the link can access but maybe that didn't work. I approved your request anyway.

Recheck my post. I added a link to an archive with several older versions. It is difficult to find older versions of XTU to download.

It's a long shot, but have you tried XTU? It is hit or miss whether it allows you to adjust it. Here is a link to download older versions that I have archive. Try one with a date close to when your CPU was released.

Yes, he does.

I got a response from the TB developer. Added to spoiler below. Corrupted TeamGroup Delta DDR5-6200 CPU-Z Report.txt Known Good Corsair Vengeance DDR5-6400 CPU-Z Report.txt

I don't like it either. I do not mind the light, but having to run crappy bloatware to change it to white really sucks. And, all of the software made for RGB is totally trash. That said, that person is just guessing. I don't know that it is accurate. If the Corsair stick that died actually died and did not get corrupted, it has no RGB. I selected it as a replacement specifically because it is not RGB memory.

Maybe this is a bigger problem than we know. Apparently, I am not the first and it is happening on other brands as well. Post #1 and #6 are of particular interest. Check this out: https://www.overclock.net/threads/corrupted-ddr5-spd-fix-tutorial.1795935/ @johnksss

I think the reasons were entirely self-serving. Not because they care, and not because they are "good people" because the exact opposite is true. They did it for self-preservation, because they know there are still multiple millions of people--including businesses--that are not going to drink their urine-flavored Kool-Aid and will abandon Windows when they can no longer use Windows 7. Some will switch to Linux, some to crApple, and some begrudgingly to Winduz 10/11 only when they are forced to because they did not plan ahead carefully enough and/or the software their business requires to function can't work on Linux or crApple. The tiny group that will be dumb enough to jump to Mac haven't figured out that Windoze 10/11 suck so bad precisely because they are reading from pages in the crApple playbook. For that reason, I probably will not upgrade my GPU any time in the foreseeable future. And, if and when I do it will be because it is supported by Linux. I am very close to the point that I don't give a rat's ass what the Redmond Retards do about anything. I value their new "products" roughly the same as I do the ABC gum stuck on the underside of a table in a booth at a filthy diner on the dirty side of the railroad tracks.

I asked because your phone screenshot says "collect in person" so I thought that meant you were picking it up.