Daharen

Lol, on the one hand, I "think" you are correct. On the other hand, I knew people who saw the trend of GPUs and CPUs getting more efficient, and figured they'd be solid with a 800 watt PSU for the foreseeable future, and then discovered that as competition rose, you start seeing some GPUs and CPUs cranking power limits hard again (With some "spiking" above 1,000 W with high end rigs despite averaging 800-850 under load). 

While this is exclusively in the range of the enthusiast space, that's the range I like building in... Sooo... Admitably, 1600 is WAY above anything I've seen other than multi-gpu rigs, and I could probably have settled at 1200 and never hit my ceiling even with the most ludicrous current setups, say an OC'd water cooled 3090 with custom PCB and OC'd 5950X (Probably wouldn't bother OCing a AMD CPU, but I'd have to at least "try" and see if I could get consistent benefit out of it, or if I just won the silicon lottery). However, with that sort of spec, and the type of overclocking I enjoy, I might have been pushing it a bit close still (Again considering spikes, which I know most PSU's can handle greater than their declared max, but some will literally shut the system down if they get spikes above their rated power, even if they could theoretically handle it unsustained). 

So... I am wary of your statement being made with confidence. It's not that I think you're wrong... This uptick in power consumption could very well be short-term and we could return to ever decreasing power ratings for the foreseeable future until my PSU is totally stupid levels of overkill, however, ya never know, and in the enthusiast space, better safe than sorry seems the better solution... I was sincerely considering a 2,000 Watt power supply that was on the market when I bought my 1,600 watt, but it would have required me to rewire my outlet and breaker, and it was Platinum rated, and I like the fact that I can run very very close to the exact power draw of the system while still having a ton of overhead because of the PSUs efficiency (At the low ranges I know it runs closer to "gold" and it's titanium rating accounts for its entire range, so I know I'm sacrificing some efficiency, but still). 

None the less, I appreciate your reply, and that it probably applies to 99% of consumers (And "PROBABLY" even me). 

11900k is much larger
https://www.youtube.com/watch?v=rBTb1tM0SDY

It's just water tubing...
 
1600w psu is not future proofing btw. Unless you have 8 gpus, you won't use it ever.

Mixing tubing isn’t an issue of any sort, it it makes your system look better and easier to assemble, then do it.

A little discussed topic that is becoming more relevant than Moore's Law in computing spaces today. After all GPU processing is all about parallelization, and CPU processing is becoming increasingly focused on core counts, and software is beginning to adapt accordingly, so a discussion of Amdahl's Law which as more to do with the growth of parallelization over time, then the growth of raw computing power over time, should become a point of discussion, don't you think?

While I can't give you a straight answer and probably no one can, it's almost certainly software side. The only way to know for sure would be to test another computer with a different SSD, and all other variables removed. 

That said, particularly with online titles in the early days of SSDs some developers realized it was possible to abuse the loading speeds to get unfair advantage over other players with your SSD, and they forced artificial load times, IE timers, instead of loading at the maximum possible speed. This would be done purely via identifying the type of hard drive, and so wouldn't effect HDD's which if this is the case, and the timers were set extremely aggressively, might result in the SSDs loading 'slower.'

If you suspect this is the case, simply check the usage on the SSD during the load screens. Set the polling to maximum, and record the data so you can go back over it in detail. You should see your SSD barely being used at all if it's just a timer artificially limiting your loading, if it's actually some sort of hardware based problem then you will see the SSD maxing out during these periods. In the latter case you're dealing with very poor programming and optimization, so still software, but of a less intentional and malicious sort, and a more incompetent and negligent sort. 

Unfortunately if I'm right there's not much you can do about it. I suppose you could run HDD's in a very big RAID 0 array to get close to SSD speeds, and still end up having the hardware register it as a HDD, but that's assuming the developers weren't clever enough to do something to check for RAID usage. No matter what if they are trying to limit your ability to load too quickly, odds are they will succeed and you'll get worse load times ironically the faster your drive (Or at least artificially set and determined load times that 'could' be worse). 

No real way for me to verify. You could write the developers and ask though, they will usually concede these sorts of things 'vaguely' so long as you don't ask for details. 

As a side note, when it comes to loading games, often times you deal with issues where the problem is more complicated than people realize. IOPS is not as intuitive as people think. I think most benchmarks use 8kb files as the IOP measurement, but it matters because latency becomes the bottleneck with more variable read and write data. If you talk to a programmer who actually has to deal with loadtime optimizations they can go on for hours about the headaches this causes, and it is in fact the reason you see SSDs that are literally in some cases 100x faster than HDDs from years ago, but the load times are maybe improved 20-30%, which is definitely not even close to proper scaling. 

Most people think IOPS is what determines this, and while they're right (Sequential only matters for large organized unprocessed data), IOPS vary based on the data size, and at really small data packet sizes you begin to become more effected by latency of the Hard Drive than by the actual throughput it can handle in general, and more importantly at REALLY low depths, talking the movement of data sections in the bits or bytes rather than Kb or KB, the IOP numbers listed in marketing materials become meaningless, your processor and memory become the keys, and in fact, you benefit more from REALLY tight memory timings than REALLY high memory speeds, and you need a processor with an outstanding IPC and extremely high single thread frequency, and even then you'll still never be able to even know if the SSD is being utilized properly or not, as it's seldom even possible to eliminate all the bottlenecks to measure the performance capacity when requests are that small, and latency is the primary factor and not throughput itself. 

This relates, because if you check you'll see HDDs have a tendency to offload 'slightly' more data to RAM than SSDs. In ALMOST all cases this has practically zero effect, and the difference is nearly completely negligible, we're talking never more than a few extra hundred megabytes of RAM usage... But a couple hundred megabytes of RAM usage when ALL of that data is nothing but small requests, can make a big difference, because the HDD isn't having to call that data at all, and now what you're really doing is comparing the speed of a partial RAM drive (Inadvertently) to a SSD. The RAM drive will always win in this battle.

It's possible but very very unlikely this is what is happening. Again this would be poor programming if that was the case. Programmers should always push all data subject to frequent small requests onto RAM, that's what it's for, but when a program is poorly optimized that doesn't happen, and programs can discriminate based on the type of hard drive because in general it can be more efficient not to hog up RAM when you can quickly pull the data off of a faster SSD, but only so long as the data being pulled is in larger request sizes that don't force the RAM and CPU overhead to increase. 

Try a linux distro to rule out hardware issues. I suspect some program left some registry entries or the like in your windows and so you might have to clean install.

In order of severity of issue:

1) Improper TIM application, make sure that the entire die is covered, and making contact with the cooler and the IHS, while generally I hate overzealous TIM application, I've never seen MAJOR issues caused by non-conductive TIM being over-used. Overheating typically occurs where too little TIM is used, so maybe add a bit more (This is very unlikely with thermal paste, and much more common with a liquid metal application, but it's certainly something that does happen).

2) Uneven or inadequate mounting pressure. While  it is possible to cause damage to a CPU by over-tightening, this is EXTREMELY rare. With direct die it's more common, but even then if you're using a proper mounting plate,  it's uncommon there. Try removing the cooler, and remount, one rotation per screw in an X pattern until finger tight to ensure even mounting pressure (This is an extremely common problem  and the most likely explanation  for 'most' overheating issues, but also annoying because it's not the cause for a significant minority of them). 
 
3) Bad pump. It's perfectly possible the fan is working, and the CPU has no issues, but if the water isn't moving, then it will quickly overheat. If the water IS moving it will still take awhile to overheat even if the fan isn't  running, because it has to saturate the thermal capacity of the water before it really begins to overheat. Usually when you see overheating, the main culprit is a bad pump more often then a slow fan, since even the radiator without a fan still does some passive heat dissipation.

4) Loop blockage. Kind of like a bad pump and with an AIO effectively just as bad. In this case your pump will run, and the fan will move, but if the AIO has grown something inside it obstructing the flow of the water/fluid, then you won't get good heat dissipation. Unfortunately, this one is impossible to test without taking apart the AIO, which voids the warranty, and is a horrible waste if it isn't the problem.

5) Stunted/Broken Fan. This doesn't always mean that the fan is not spinning, it could just mean it's spinning so slowly that it's not moving a significant amount of air through the radiator. If this is the case then you shouldn't IMMEDIATELY hit high temperatures if you turn the PC on from cold and start running a large load. It will take awhile to saturate the water in the  AIO, so this is a good test to see if this is the problem, if you do immediately hit high temperatures, it's something else... But if your temperatures just gradually climb and your AIO isn't keeping up, your culprit may be a shotty fan not helping the radiator cool fast enough. 

6) IHS issues/Stock Thermal Paste... I am hesitant to post this one, but I have seen bad thermal application within an IHS before, and while I don't generally advise delidding, if this is the case, some of your thermal problems could come from this. It's rare, but it's common enough (Less than one in a thousand but it happens), that it's worth mentioning. 

7) Now we're on the fringe cases, but not impossible. Some IHS's (Or in the case of direct die cooling, CPU dies) are not within the same level of tolerances as others. USUALLY this is something that would be caught by QA before the product gets to you, but things slip through the cracks sometimes. An uneven IHS, or uneven CPU die could result in the die having poor contact with the cooler or IHS, or the IHS having poor contact with the cooler. The only way to test this really is to get a micrometer and ask some other people with the same CPU and a micrometer of their own to compare measurements with. Assuming your IHS is within normal tolerances, then you have to delid to test the die... I stress this is VERY unlikely, but again not impossible. 

? If your cooler is working, and the CPU is within spec, and you're still getting temperatures like that with a proper thermal paste application and cooler mounting, then you're starting to enter the territory of borderline impossible errors. There are so many safety features on most motherboards and CPUs that pumping in the power necessary to cause that sort of heating without frying the CPU, or tripping one of the countless safety features on the Mobo is nearly impossible, and will almost always result in shut-down... However, I guess while at this point I couldn't give you a solid answer, if it's none of the above, you might want to check the voltages actually being delivered to your CPU by comparing the CPU sensors to the motherboard sensors with HWinfo... If you see there is a gross difference between the two values for your V-Core or CPUs integrated graphics, then you might have some one in a million fluke where the power delivery is screwed up and causing thermal issues... Power delivery problems like this usually have to get past safety features on the power supply, mobo, and CPU, so it's exceedingly unlikely... But as with the others I suppose not impossible. 

In order of severity of issue:

1) Improper TIM application, make sure that the entire die is covered, and making contact with the cooler and the IHS, while generally I hate overzealous TIM application, I've never seen MAJOR issues caused by non-conductive TIM being over-used. Overheating typically occurs where too little TIM is used, so maybe add a bit more (This is very unlikely with thermal paste, and much more common with a liquid metal application, but it's certainly something that does happen).

2) Uneven or inadequate mounting pressure. While  it is possible to cause damage to a CPU by over-tightening, this is EXTREMELY rare. With direct die it's more common, but even then if you're using a proper mounting plate,  it's uncommon there. Try removing the cooler, and remount, one rotation per screw in an X pattern until finger tight to ensure even mounting pressure (This is an extremely common problem  and the most likely explanation  for 'most' overheating issues, but also annoying because it's not the cause for a significant minority of them). 
 
3) Bad pump. It's perfectly possible the fan is working, and the CPU has no issues, but if the water isn't moving, then it will quickly overheat. If the water IS moving it will still take awhile to overheat even if the fan isn't  running, because it has to saturate the thermal capacity of the water before it really begins to overheat. Usually when you see overheating, the main culprit is a bad pump more often then a slow fan, since even the radiator without a fan still does some passive heat dissipation.

4) Loop blockage. Kind of like a bad pump and with an AIO effectively just as bad. In this case your pump will run, and the fan will move, but if the AIO has grown something inside it obstructing the flow of the water/fluid, then you won't get good heat dissipation. Unfortunately, this one is impossible to test without taking apart the AIO, which voids the warranty, and is a horrible waste if it isn't the problem.

5) Stunted/Broken Fan. This doesn't always mean that the fan is not spinning, it could just mean it's spinning so slowly that it's not moving a significant amount of air through the radiator. If this is the case then you shouldn't IMMEDIATELY hit high temperatures if you turn the PC on from cold and start running a large load. It will take awhile to saturate the water in the  AIO, so this is a good test to see if this is the problem, if you do immediately hit high temperatures, it's something else... But if your temperatures just gradually climb and your AIO isn't keeping up, your culprit may be a shotty fan not helping the radiator cool fast enough. 

6) IHS issues/Stock Thermal Paste... I am hesitant to post this one, but I have seen bad thermal application within an IHS before, and while I don't generally advise delidding, if this is the case, some of your thermal problems could come from this. It's rare, but it's common enough (Less than one in a thousand but it happens), that it's worth mentioning. 

7) Now we're on the fringe cases, but not impossible. Some IHS's (Or in the case of direct die cooling, CPU dies) are not within the same level of tolerances as others. USUALLY this is something that would be caught by QA before the product gets to you, but things slip through the cracks sometimes. An uneven IHS, or uneven CPU die could result in the die having poor contact with the cooler or IHS, or the IHS having poor contact with the cooler. The only way to test this really is to get a micrometer and ask some other people with the same CPU and a micrometer of their own to compare measurements with. Assuming your IHS is within normal tolerances, then you have to delid to test the die... I stress this is VERY unlikely, but again not impossible. 

? If your cooler is working, and the CPU is within spec, and you're still getting temperatures like that with a proper thermal paste application and cooler mounting, then you're starting to enter the territory of borderline impossible errors. There are so many safety features on most motherboards and CPUs that pumping in the power necessary to cause that sort of heating without frying the CPU, or tripping one of the countless safety features on the Mobo is nearly impossible, and will almost always result in shut-down... However, I guess while at this point I couldn't give you a solid answer, if it's none of the above, you might want to check the voltages actually being delivered to your CPU by comparing the CPU sensors to the motherboard sensors with HWinfo... If you see there is a gross difference between the two values for your V-Core or CPUs integrated graphics, then you might have some one in a million fluke where the power delivery is screwed up and causing thermal issues... Power delivery problems like this usually have to get past safety features on the power supply, mobo, and CPU, so it's exceedingly unlikely... But as with the others I suppose not impossible. 

It still seems to be working as of 5 months ago. https://www.youtube.com/watch?v=tnd2LO0IBic

 
I'm guessing you'd replace the cpu before it becomes an actual problem. It would depend on how much of the original coating is present.

The protective layer including more diamond in it may increase thermal conductivity but would probably be less reliable. https://aip.scitation.org/doi/pdf/10.1063/1.4978043
I'm not sure what the cost would be but it's not within your budget nor practical. It's a process that's most likely limited to custom orders of silicon wafers and the process for getting it on a specific cpu would be proprietary to its manufacturer. 

IANAL:
 
I doubt a PC would be seized for exactly the reason you stated above: reselling it would be more trouble than it's worth. Creditors/trustees will want the stuff with relatively universal value.
 
As far as valuation, you're right. You've voided basically every warranty. You've damaged the CPU to the point where its value is tremendously depleted. Your GPU has no warranty and has been altered in a way that could have caused damage. The easiest way to do this would be to go back to the value of the parts when you originally purchased the item, then factor in the damage caused by your modifications. Remember that these components have essentially zero value outside of this system at this point because of the extensive modifications you've made to them, so their individual value is going to be largely based on the system as a whole.
 
Now, go back to the MACRS table in the five-year, 200% class (table A-1), which is what you'd use if you were claiming a tax deduction on your PC, so that seems fair to use here. The MACRS depreciation table is as follows:
Year 1 (year of purchase): 20% of original purchase value lost Year 2: 32% of Y1 value lost Year 3: 19.20% of Y2 value lost Essentially, if you're more than one year but less than two years removed on a PC purchase, roughly 50% of its value is gone. Yes, I used to deduct PCs on income tax. Yes, it sucks to calculate, but I never got into trouble nor was I audited using this table. Don't list the components, just list the overall value based upon original item price minus the value lost by modifications, then depreciate based upon the table above. So, example, if your PC was worth $1.5K when bought and you drop that to $1K based upon modifications made, and you bought that PC more than two years ago but less than three:
Original PC value: $1,000 Year 1 PC value: $800 Year 2 PC value and your current value based upon MACRS: $544 Again, IANAMFL, but when I was calculating PC value for my own tax returns and for my old company's tax returns, that MACRS table is how I did it.

That makes sense, and I'll  just look up one of the all in one watercooling kits and apply 60% to that as well to get a rough estimate... I also have noctua fans so I'll account for those as a replacement. 

Will still come out quite valuable, but hopefully below the margin they'll aim for liquidating. 

Not a good thing. Maybe you will get lucky.

Yep, I'd also agree with you by the way, except I've remounted it and reapplied now a total of 3 times with the same result in the same spot. Each time maximizing the pressure just short of using tools to increase mounting pressure (Though it's so thumb tightened I need to use something to unscrew the mount often for reapplication). I doubt it's mounting pressure given that fact, and I'm fairly confident in saying that if a bubble was appearing consistently in the same spot that would not likely be application error, but something to do with the shape between the two surfaces, as I can't imagine being that consistently unlucky. 

With paste, it doesn't happen, I get higher average temps, around 30c, and max out around 50-60c under maximum load, and there is no significant core to core variation. The issue only occurs with liquid metal. I am applying to both the block and the die, and getting full coverage with no pooling. In one instance (This was a 4th application I'm not counting in the three mentioned above), I tried a more generous application of liquid metal, but as expected that didn't have good general results, as liquid metal really doesn't gap fill even when overapplied, tending to eventually settle or be pressed out from between the mounting surfaces. 

Anyway, it may not be height variation, but it's the only thing I haven't really tried. I was on the forums a week or so ago about this problem, and have been troubleshooting since then, this is the last step and if this doesn't work I really have no clue what is going on and might just have to switch back to paste since I know that works at least. 

The clock speed at which the CPU operates in is controlled entirely by the CPU. The OS can request it to go to certain power states, but this does not directly control how fast the CPU operates or how many cores are being pushed.

That's how CPUs work.
When they need to compute something the frequency boosts up.
Even at idle there are thousands of things happening in the background of a computer.

If you have no programs running then a clean OS might downclock and stay there for a while, or on low power modes that prevent boosting.
No there is no harm done other than slightly higher temps.

The silicon diffusion barrier on most CPU's is silicon nitride. You cannot redo this coating if you lap the die down. I don't really know if liquid metal can diffuse into silicon readily over time, it might be a case of it takes 10+ years, or it could be a case of "Well I got two years out of this thing at 1.4v 5.4GHz so I guess it was either this or it degrades in the same time span". 

You pretty much answered your own question as to why it's a meme... simply put, it's because in 99.9% of builds it's not needed/warranted to watercool your RAM modules. Most time just the heatspreaders and a fairly good airflow over the ram would keep it cool, even overclocked a fair deal.
With DDR5 on the horizon too, only people with money to burn, and/or that absoloutely NEEDS faster ram to help with their workload would buy the expensive faster ram, or watercool it right now IMO. As you said, there's way better things to spend your hard earned cash on. For most people faster than 3200 ram isn't needed, as their use case might not see any benefit, or will be negligible.

What are you planning that will require 2 or more pumps in series, I'm assuming this is due to a requirement for head pressure? 
 
To answer the question you can use a power hub to have a singular PWM control line come in a group control all your pumps so they run at the same speed. Something like this from Swiftech will work well.
http://www.swiftech.com/8-WayPWMsplitter-sata.aspx
 
As for pump option if you just require a large amount of head pressure I'd suggest to just get one of these as it will have a substantial performance over combination of DDC's or D5's.
https://koolance.com/pump-g-1-4-bsp-pmp-600

I use generic 1ml syringes fitted with a fine plastic tip to manipulate liquid metal. Haven't tried metal tips. I find it good for producing a small drop of LM where I want it.
 
Can't give a general source for the plastic tips. I got mine as part of Salifert water testing kits. It is similar size point to that provided by the Conductonaut kit.
 
Image of the syringe I use at end of thread below:
 

I had a similar issue with the rear fan, i switch in bios from PWM to DC.

Do you mean in DC mode?
Absolutely!