Mr. Fox

I'm happy to hear that you are making progress but sad that you are having to go through it.

The thing that is even more stupid than Samsung doing this is the reality that some people will fall in line, like good little sheeple, and do it. The will rationalize the decision because they "have nothing to hide" or because it is "for the greater good" but at the end of the day everyone else knows they are idiots that are a danger to themselves and everyone around them.

He is using the G.SKILL Trident Neo 6000 C26 2*16 1.450V magic sticks. It is unusual to find any others that can do so well, and part of the reason for that is the effort ASUS has put into making them be able to.

@tps3443 did you see this? https://www.overclock.net/posts/29599251/

It is impressive how close to the same performance the 270K Plus delivers compared with the 9950X with 25% fewer threads to handle the workload. The 270K Plus is actually just a tad better, and it's half the cost. Not to even mention the fact that the 270K Plus is much easier to tune and runs circles around AMD's best where memory overclocking is concerned. 💖

https://hwbot.org/benchmarks/cinebench_-_r23_multi_core_with_benchmate/submissions/6046141 With E-cores bumped to 54x... https://hwbot.org/benchmarks/cinebench_-_r23_multi_core_with_benchmate/submissions/6046144

Wow, down to just under 60% now. Wonder why it is still so high? Micro$lop deserves to lose another 25-30% market share, but that might not even be enough loss for those idiots to pull their head out.

I've never truly liked them and was always a desktop-only guy until a career change had me only home only on PTO, holidays and weekends for about 12-14 years. That's why I accepted the big Alienware 17 and 18 inch and then Clevo DTRs as a necessary evil. After a while I kind of grew numb to how pathetic they are and even began to view them favorably. I even made lame excuses and rationalized a degree of legitimacy for their glaring shortcomings, but it did not take long after moving back to desktops for the misguided novelty to die and my disdain for them quickly reignited. My new job is going to require a little bit of travel, unfortunately. My last two business trips I did not even bother packing my Precision turdbook. If something happened to it, there is a fair chance that I would probably not purchase a replacement turdbook. Using a notebook keyboard and screen is annoying to me, even for web browsing. The only thing worse to me than a turdbook is a slimy touchscreen smartphone or tablet, but all of them are garbage cut from the same cloth to me now. Turdbooks are engineered from end-to-end as compromised products, in varying degrees of trashification... none actually worth a damn. I really like my 5080 Prime. I might end up with another 5080 at some point. The thought of selling my 5090 for significantly more than what I paid for it, spending 25% of the proceeds on a 5080 and doing something else with the remaining 75% has crossed my mind a time or two. The possibility for regret is the only thing that gives me pause. I'd not consider buying another 5090 again but I own it and it is paid for, so why tempt fate? Finally stopped tinkering with the new hardware toys and took a few minutes (literally all it takes) to get Kubuntu installed and configured on the Z890 Apex. I'm so used to it now that a clean Linux install is faster and requires less effort than debloating, removing garbage and the extensive manual de-configuring of all of the trash necessary to make Windoze an acceptable OS. Having to use the Windoze Cancer OS for everything was starting to make me feel kind of icky on the inside. The longer I use Linux, the harder I find it to forgive the Micro$lop clown posse for their atrocities. The only thing saving their lousy hind ends from perma-death is the Windoze-only software that I sometimes want or need to use.

Easy and hassle-free installation on the Z890 Apex. Everything just works, and works well. No debloating or tweaks necessary.

If the BIOS mod was created by Prema after the GPU was released it should be supported and it is about 99% certain it will be supported. If it was before the GPU was released then it is very unlikely. I'm glad that I don't have to worry about that kind of nonsense anymore. Sometimes I forget how bad laptops suck. Even the better non-BGA turds with desktop CPUs and MXM suck without a lot of stupid rigmarole. In hindsight I don't know how I put up with them for as long as I did.

Actually this version might be even better. I adjusted more obscure timings after doing a bit more research. Totally stable with my G.SKILL Trident F5-8400J4052G24G kit. Testing now with the TeamGroup Xtreem CUDIMM kit and seems to be identical. Below is a link for the settings. Z890 Apex 8600 Validated - BIOS 9901

Does your chassis have fans in the bottom blowing directly on the cable and GPU? I also wonder if the width and frequency of the temperature delta between room temperature (cold) and maximum sustained load temperature affects anything, even when the high is totally fine, how often it experiences that range from low to high to low. The frequency of that delta shift might be causing fatigue of these incredibly fragile turds? It is unusual how fast your system chews through these crappy cables. But as you said, thank God for the WVP2. It is a reasonable expectation for PSU cables to last beyond the lifespan of anything in the system, including the PSU. But, these are such $hitty examples of engineering incompetence that it is not the case. There is no reason that anyone should ever have to replace any PSU cable due to any reason except for accidental physical damage (cuts, pinches, etc.). "Normal wear and tear" is an unacceptable reason, especially if they are less than 10 years old.

I wonder if ambient conditions such as humidity come into play with the connection integrity? I don't think I've ever seen or heard that mentioned anywhere and I can't help but wonder if it has any effect.

Sorry to hear about the 2080. You might be able to see if Tony at Northwest Repair could fix it for you. Worth a shot asking. Probably much cheaper than a replacement or upgrade if he can. I suspect the problem is failed memory chip(s). I think Rockitcool are still in business... https://rockitcool.myshopify.com/

The developer has a thread at overclock.net and you can download it from the opening post (or his web site). It is a great program. You can export your memory settings as an XML from HWiNFO64 and import it, click the analyze button, and it will tell you want timings are non-compliant with JEDEC. If you run the stability test if it can determine what setting is off it will identify it for you. Unless the price is crazy, I plan to buy it when the free beta access expires. [BETA] Memory Foundry Pro – RAM Stability Tester & Benchmark | Free Beta Access Now all we need is a clone of ZenTimings that works for Intel. The ASUS MemTweakIt and ASROCK Timing Configurator crapware tools are buggy junk. The ASUS tool is missing too much stuff and the AROCK tool misreports some values.

I hate Meta/Facepoot and only want the worst for them. However, I am not sure I agree with them being accountable for the irresponsible behavior, poor judgment, etc. of people that use the platform. I would blame the irresponsible parents and schools for not having control of their children and raising them the right way. Blaming Meta seems as misguided as blaming guns for crimes commited with guns, blaming the keyboard for misspelled words and silverware for the obesity problem some people have. I look at this as a state cash grab/robinhood mentality by states that are failing financially due to their idiotic, incompetent leadership. Three of the four states are run by overtly leftist imbeciles (CA, CO and NJ).

Nice "Country Metal" tune dropped 2 hours ago.

So, did a little bit of digging for elusive information online and found a solution that works with my TeamGroup CUDIMM kit. Forcing the clock driver to work for manual tuning without using an XMP profile allows me to keep the NGU, Ring and D2D values the same and still boot at 9000+ (example in above post). It will not work with my G.SKILL 8400 kit (non-CUDIMM). With the CUDIMMs installed after POST there is a new menu under the Skew Control menu in memory configuration. CKD Configuration and manually set Single PLL mode forces the CUDIMM clock driver to take over and do the heavy lifting on manual tuning. If you have not booted at least once with a CUDIMM kit installed or have normal DIMM installed the CKD Configuration menu does not exist.

Definitely not a daily driver like this, but worth a few giggles... https://hwbot.org/benchmarks/memory_frequency/submissions/6044182 https://valid.x86.fr/qiv4lc https://hwbot.org/benchmarks/aida64_-_memory_read/submissions/6044183

Thanks. Happy to share it. Let me know when you get home and are ready to use it. I've got an 8600 and and 8800 CMO I can send. Edit: Here is a zip of the folder (BIOS, CMO and Settings TXT files). @Talon

I think this is going to be good enough for now.

https://hwbot.org/benchmarks/cinebench_-_r23_multi_core_with_benchmate/submissions/6043707

OK, I got it to boot and validate 9200 in Gear 4 without overclocking the CPU, NGU and Ring, so I will see what happens if I start tweaking those. Will first see if it will boot in Gear 2. If not, then maybe no point in experimenting further except for memory frequency submission on HWBOT. https://hwbot.org/benchmarks/memory_frequency/submissions/6043616 https://valid.x86.fr/kdh1u2 Edit: Nope on G2 above 8800. Back to G2 at 8800 for now. These sticks are not on water yet, so I can't run this too long due to thermals. It doesn't take long for the sticks to hit 50+ °C running Karhu on air. (About 3 minutes.)