Skill DDR5 2x16GB 7800 and the other is 2x24GB 48GB 7200. Both of these kits have passed Memtest86 so I know the RAM isn't the issue. Had a few issues with games not liking XMP speeds so I have just left XMP off lately to eliminate those issues.
These are kits with a highly enthusiast-grade XMP, this kind of XMP is rarely plug&play, also see the explanations
here. If you don't have a top-notch IMC (integrated memory controller of the individual CPU), then most likely this XMP will need manual tweaking, GODLIKE board or not. Even the GODLIKE board is not ideal for such DDR5 speeds, because it has two RAM slots too many. In the past, they had the UNIFY-X boards with only two RAM slots, which is electrically much more ideal than having an empty slot per channel before each of the modules. Since most people should just stick to using two DDR5 modules anyway, and having only two DIMM slots results in superior signal quality over four slots with two empty ones. Sadly those two-DIMM-slot boards have become rare, even though four modules tend to run quite badly. The RAM itself can do the speed in isolation (it was binned for it after all), however, the memory system also involves the CPU's IMC, the mainboard and its BIOS. See my
RAM thread for further explanations.
Most of the time, the IMC will be the limiting factor at these kind of speeds, it needs manual tuning, maybe XMP cannot be stabilized at all (because your kits are so high-end that the IMC needs to be top-notch to handle it). And even the signal quality on the board becomes questionable in the high DDR5-7xxx range. In those cases, what you could try is to enable XMP, but before pressing F10 to save&exit, you also set "DRAM Frequency" to something lower than XMP by hand. So for the DDR5-7800 kit you could try DDR5-7200 or -6800 instead. This will take a lot of stress off the memory system because it doesn't have to run at the absolute limit of the technology or beyond.
It's good that you verified that none of the kits are physically defective (meaning, throwing errors even at the safe JEDEC speeds) and that the IMC is ok in principle. It probably just all goes down the drain once you enable this extremely demanding XMP.
Had a few issues with games not liking XMP speeds so I have just left XMP off lately to eliminate those issues.
Again, this would be a huge waste of money for those nice kits, because you effectively demote them to cheap standard kits you'd get for a fraction of their price. So the better approach is to find out what your memory system can actually handle. Of course, we have an underlying issue of CPU-related instability, so on the 14900KS this wouldn't make sense now, but on the stable CPU it would definitely be the way to go.
PSU is Cooler Master MWE Gold 1250v2. Hard drive is M.2 WD Black 4TB.
Wow, the PSU is the first mid-range model i see of anything. Nice mid-range, no reason to worry at all, but still. SSD is a good one, i have an SN850X 4TB as well. One of the best PCIe 4.0 x4 SSDs, and thus one of the best overall, because the PCIe 5.0 ones are not very good so far (they run way too hot, they need to come out with controllers with way lower power draw for them).
Temp sitting in BIOS is usually 30's. Windows with a couple browsers open is usually low 40's. Seems well within proper operating temps and I wouldn't expect to break any coolness records with a Cooler Master 360mm AIO.
Ok, in the 30s in the BIOS, that's more like it. So most likely you have the browsers creating some background load. 360mm AIO is a good choice for any -KS CPU, if it's a well-performing AIO. In fact it should be the minimum for anyone with such a CPU, because simply put, there isn't anything remotely like a -KS when it comes to the power draw. So ideally you'd use the biggest and baddest AIO that fits your case.
We could check the sensors with
HWinfo64. Run it and open "Sensors", then expand all sensors by clicking on the little <--> arrows on the bottom, also expand the columns of the sensors a bit so everything can be read. Make it three big columns of sensors (or four, if the screen resolution is high enough). In the end, it should be a screenshot with all the sensors visible at once, like this:
Make sure the power plan in Windows is on "Balanced". Do nothing on the PC for a while (couple minutes), so the "minimum" baselines for the values are established. After that time in idle, then produce full CPU load with
Cinebench R24, and after completing a 10 minute run, when the CPU temperatures have stabilized at the highest level, take a screenshot of the sensor window and tell me the Cinebench score. On this screenshot, we can see how it runs with fully multithreaded CPU load, the highest normal load you can encounter in daily use without resorting to an artificial stress testing tool like Prime95.
If the CPU causes a BSOD early in Cinebench, try to run it for as long as you can and then take a screenshot before it BSODs. If we can catch a glimpse of what it does under load, that would probably give some hints.
I know the 14900KS is a side-grade but from what I have seen the 14900 CPU's run faster and cooler than the 13900 CPU's (although KS might be different). So if I can drop the temps a little and get some extra performance I am all for it. I also don't mind having a spare part laying around in case I need it, I help my friends build PC's too.
14900KS always runs hotter than a 13900KS. The manufacturing process has not changed, yet they squeeze another 200 MHz peak out of it, so on average it will need even higher VCore than the 13900KS already did (although the core voltage the CPU requests is individual for each CPU, hence, on average). Every review i saw of the 14900KS, comparing it to the 13900KS, the 14900KS has broken new power draw records for a desktop CPU. And higher power draw directly translates to higher temperatures.
Was hoping to hear from some folks that tried the 14900KS with a MSI z790 motherboard to see if it even worked for them. I have proven my system is stable with a 13900KS so I am wondering if it is the BIOS or the 14900KS. I will RMA this CPU if necessary if I can find out if the BIOS isn't the issue.
I doubt it's the BIOS, because then everyone with a 14900KS would have this problem. All the 600/700-series boards use the same BIOS codebase, you can see all the BIOS changelogs are similar, if not identical across all those board models. So if they mess something up for one CPU model, everyone would be affected. That's why it points much more to your individual CPU than anything else.
