Mr. Fox

Yes, in fact it equaled my custom loop and I started trying to figure out why. I was using the 13700K (not delidded) and the temps were a couple of degrees cooler with the EK AIO. I took apart the Optimus block and found debris in the fins and cleaned them out, and after that the custom loop was just barely better. So little better, in fact, than one could say the added cost and effort is not worth it. The custom loop has benefits like inclusion of the RAM and GPU blocks. But, for a person that plans to leave those cooled on air, the AIO is the smart way to go financially. Maybe not just any AIO, but this one at least.

I look forward to seeing that. White or otherwise, I'm sure it will be nice. That EK 360 AIO then I'm going to be publishing a review on here shortly does an impressive job. Much better than I expected. You can get a TeamGroup 4TB NVMe for $159. It is not as fast as the SN850X, but for $100 less it doesn't need to be. It'll get the job done and that price is just really too good to pass up.

Speaking of temperatures. Pulling cold air through a radiator is always better than exhausting hot air through the radiator. And, with the top radiator blowing on the RAM sticks, it runs cooler, too.

With a TDP-handling capacity the functional equivalent of this and a thin chassis suffocating what little resides inside of it... probably so.

The really great news (for me) is that because it is a laptop product it just doesn't matter to me how it turns out, LOL. That it will be a pile of manure is taken for granted before any details are known. 😉

That is a great picture of both of you. Thanks for sharing it.

I hope this fixes my Micro$lop Office text rendering problem. Still fighting this nonsense with the 6900 XT. It only seems to affect Word and Outlook, but it is extremely annoying. It is very difficult to concentrate and not feel disoriented when part of the information you entered in an email or a document 1 second ago randomly disappears (becomes invisible) momentarily. Drives me nuts some days when it is really bad. There are other days that it never happens. That's good news. It will make the ludicrous pricing of the 4090 a little less painful if I can get 2 years out of it before it becomes obsolete. What is the temperature of the DIMMs when it starts erroring out? It may be thermally induced. I can overclock my memory much higher without error when liquid cooled than I can trying to use a fan blowing on the stock heating blanket trash that memory ships with. The chance of having errors and random BSOD issues is much greater at 45°C and higher. Same was true of DDR4. The relevancy of memory stress tests like TM5 and MemTest Pro is diminished when errors are occurring due to heat. In real-world utilization you may never encounter any instability issues at all like you do when stressing the crap out your RAM and overheating the modules using a memory stress test. Probably the more reliable way to test is with benchmarks and extended gaming sessions. The memory modules generally do not get nearly as heated up doing things that are normal use scenarios. The load some of these memory stress test produce is not a realistic representation of what would occur in typical use, even the more stressful of the typical use scenarios.

I actually had some moron on eBay offer me $250 for the Z690 Apex, DDR5-7200 and Kryos block bundle. I responded to his message with nothing but "LoL" and left it at that. I'll let it rot and collect dust on a shelf in my office rather than sell it for that price. Had he offered that for just the motherboard only I might have considered it. I think it was a driver problem. I did a lot of aggressive registry editing after running DDU and manually deleted everything I could find in the registry that made reference to "Radeon" and "6900 XT" and "6950 XT" and reinstalled drivers after rebooting. So, now it functions correctly again. These are basically the same GPU and I don't know why they created a clone of the first. I guess they just needed something with the different name to sell. Some drivers can't tell which one is installed because they seem to only differ in firmware and hardware ID so it gets named "6950 XT Series" which doesn't even make sense. I believe that running DDU between driver installs was not good enough and leaving garbage behind that was causing conflicts. Just a guess on my part, but I don't have any other explanation for it. Seems logical based upon what fixed it. This was a weird flashback to like 10 or 12 years ago. I remember back then having similar goofiness with drivers not being able to identify whether the GPU was Radeon HD 6970 or HD 6990 and assigning a generic "6900 Series" name in Windows and that caused random glitches in behavior. It was also annoying because when you uploaded your benchmark scores to 3DMark it identified the wrong video card most of the time. Buggy firmware, drivers and software have always been their primary Achilles heel.

Yes, wait for 14th Gen and exploit the opportunity to buy the best used 13900KS you can find for a fraction of the scalper prices silly people originally paid for them. 14900K is a refresh of 13900K/KS with a 200MHz turbo clock bump.

Bundle is still available. If you want only the motherboard or waterblock send me a PM to talk price. RAM kit is not available for sale without the motherboard, but motherboard alone (without RAM) or water block is negotiable. (RAM will be available if motherboard sells without it first.) My thought process here is someone upgrading from a DDR4 system to newer gen can benefit from having the DDR5 with the motherboard. If you don't want the waterblock I will knock $100 off the price of the mobo and RAM. The motherboard and RAM are sold. Nobody was interested in the waterblock. I didn't realize how few PC enthusiasts were left in the world.

Ghost Spectre is a nice/fancy Winduhz 11 mod for people that value style over substance. I think it's the most pleasing of any Winduhz 11 experience, but not the best performance. It carries too much bloat and only offers a minor, and I do mean very minor, bump over a full-feces OS.

I'd have to test it again to know for certain. Most versions of Windows 11 22H2 (bloated or debloated) seem to perform better in Port Royal and it is mostly a CPU agnostic benchmark. Port Royal is probably the one benchmark I run the least in 3DMark suite. I don't like it. I am also not a huge fan of Speed Way, but I like it better than Port Royal. Sky Diver and Time Spy are my favorites in the 3DMark suite, and they are both heavily influenced by CPU/Physics performance. Fire Strike as well, but it doesn't give as much weight in the overall score for CPU/Physics performance as I would like it to. It does give decent weighting to the Combined test.

As @Papusanmentioned, the only way to know which OS works best is to test them. Ghost Spectre has a nice GUI and features, but it is generally terrible for CPU performance (Cinebench, 3DMark Physics and Combined tests, etc.) but good on graphics-only tests like Port Royal and Speed Way. The overall best OS for CPU performance I have found (excluding Windows 7, which is superior to everything released since for CPU and memory performance) is Oprekin.com Lite Plus W11 Pro 21H2. Almost all versions of Windows 8, 10 and 11 are horrible at wPrime32M benchmarks. They take roughly twice as long to complete wPrime32M compared to Windows 7. Sky Diver, 3DMark 11 and Vantage physics tests are also generally vastly superior using Windows 7. You can actually see the lag at the start of some of these tests. It is like nothing happens for the first second or so. Cinebench R23 cannot function on Windows 7. If you remember, it was originally released as a "Micro$oft Store-only" benchmark and they caught a lot of flak for being stupid. Cinebench R23 relies on WinRT files needed for ARM support for some stupid reason. Maybe they knew the superior EOL OS would rip it to shreds so they made it non-functional to keep the EOL OS from making the newer ones look like the trash that they actually are. As best I can tell, the really fast CPU performance tests that require a nearly immediate response as soon as it launches (wPrime32M and 3DMark 11, for example) excel on Windows 7 because of some kind of lag on later releases. Maybe some kind of aggressive power management thing. Longer tests, like Cinebench generally do fine on Windows 10 and 11. A few benchmarks, like Port Royal, get the highest scores for me on the full cancer versions of Windows 11. Maybe DX12 optimization is the reason. I am not sure of the mechanics behind any of this. All based on scores and observations, not technical information. That is why the only way to know is have multiple OSes installed and go to the trouble of testing.

Trashmasters never get better.

Yes, same problem on W10 LTSC 2019, W10 LTSC 2021 and the various versions of W11. I have not tried an older version of 3DMark yet, but I will when I have more time to experiment. The problem is consistent enough that it "feels like" 3DMark has a framerate limiter on the 6900 XT. I often wear a tin foil hat, but in this case the performance is so low that it would be evidence of malfunction more than manipulation. (No, I would not be surprised at manipulation of results for a nefarious purpose.) If AMD were doing something crazy like paying UL to supress 6900 XT scores to sell new GPUs they would at least want the stock performance to be consistent to avoid raising suspicions of monkey business or foul play. (No, I would not be surprised to learn that AMD or NVIDIA did something like that. I would only be surprised by the mistake in not making it harder to spot the problem)

I haven't looked for reports from other users. Just too busy to make that a priority since everything else seems to be working correctly. I noticed it with overclocking, since there is no point in running benchmarks stock if you're doing that for fun versus diagnostics. After seeing it I did test stock and the issue is the same. The clocks hold, power draw is right and no throttling or thermal issues, just abnormally low scores. That's why I think the problem is 3DMark or Windows, not the GPU. I did try older drivers and the latest, and multiple Windows installations, and no change.

I'll have to play around and see if I can find what's up with the 6900 XT super low 3DMark benchmark scores. It seems to perform normally with anything else so I don't know if it's something that UL or Micro$lop did. And this surfaced a little over a month ago. I had my chiller fired up and everything nice and cold, geared up for a benching session and ended up abandoning the effort. I tried it a few times since then without chilled water and it's the same. Games play normally and all that, just 3DMark is messed up. I tested on multiple versions of Windows 10/11 as well. I tried running Speedway the other day and my score was in the 3000s and Firestrike is about 2,500 points lower than normal.

Based on my collective experiences with the brand, and especially my last one with 5950X, it is not likely I will give them any more chances to change my mind about them. I don't think they deserve my money or my forgiveness. And, you don't see me benching the 6900 XT any more because for some reason it has benchmark scores close to half what it used to and I have no idea why. Nothing seems messed up otherwise. Crazy, and it irks the crap out of me.

Speaking of using the EVGA machine for work... spent all weekend doing the EK AIO testing for the review and building this for review. Very well-made. Was a terrible pain in the butt doing the mods for my external loop. It was sobering to be reminded of why I hate air cooling. Glad it is over.

With ASUS you really should wait because I don't think they actually test anything. They toss it out for the zombies that worship at the Newer is Better Church of Lucifer to test and see if anything blows up. I think I threw up in my mouth a little bit with that one. Disgusting. It's obvious who their target market is, and it sure as hell aint us, LOL. But, yeah... that looks like doggie dung to me.

Beta firmware has never really mattered to me as long as it works and does what I want it to do. To me it is a meaningless designation assigned to it that may have absolutely nothing different than the final version other than the version number displayed. Kind of like WHQL certification on drivers. The beta, non-certified and WHQL versions can be 100% identical, with the only difference being the Redmond Reprobates have been paid a fee to put their utterly worthless seal of approval on it. Totally irrelevant IMHO.

It's kind of like the old saying, "Those that can, do. Those that can't, teach." Popularity is relative. It's probably more accurate to say "gaming is more common" because it's not more popular with those who enjoy benching more than gaming. You can certain enjoy both, as you do. Some people don't enjoy benching because it is just not their thing. If they are not very good at it, or their anemic hardware sucks, or they don't know how to tune it properly, chances are they will never like benching. I know I don't enjoy doing things that I don't know how to do, or things that I suck at doing.

@Papusanis suggesting limiting your memory capacity to 32GB to test the game to see if any different than 64GB. After test run the second command to make all 64GB available again.

Pretty much everything in the tech industry is moving to the point that it boils down to "How much are you willing to put up with and at what price tipping point are you willing to tolerate that much crap?" I am having to re-evaluate a lot of things, including whether or not being a PC enthusiast is even a thing worth pursuing any more. It is starting to feel like the term "PC enthusiast" is an oxymoron. At some point I could see PC ownership being reduced to the equivalent of owning a hammer or a set of screwdrivers. You use them when you need to drive a nail or a screw. The rest of the time they are in your toolbox. You don't think about them. You don't upgrade them. You don't show them to your friends, make YouTube videos about them, or feel a sense of accomplishment or pleasure from owning them. You might derive some level of satisfaction from what you use them to fix or create, but they are not an object of focus. If awesome doesn't matter any more, then the only questions that remain relate to purpose and adequacy. Can I make do with something from Harbor Freight, Walmart, maybe even the Dollar Store? Or, do I need to spend an absurd amount for a tool made by Snap-On that only gets used occasionally? Can I use a nail instead of a screw? If yes, then why do I need a screwdriver? A hammer is good enough. How many nails do I need to set? One or two? I think I can use that rock in my backyard for that. Why waste money on a hammer?