Question Has anyone seen a post about the best AIO Water CoolerIs for I9-9900k

Dec 12, 2018
26
0
30
Google is no help, so i cant find anyone who has tested multiple AIO's and a custom Loop on a I9-9900k OC/Non-OC to see whats best for thermals and how to achieve good thermals while still busting a good OC.

Current Build:
Fractal Meshify S2
Kracken X72 (3 Pull fan setup) (changing to EVGA CLC 280 this weekend)
I9-9900k (Liquid metal between the IHS and the AIO plate)
36Gz G.Skill TridentZ RGB 3200 (changing to non RGB version, due to issues with the RGB controller, this weekend)
EVGA FTW3 RTX 2080TI (Air Cooled)
Samsung 970 Pro 1TB
Samsung 860 Pro 1TB
ASUS ROG MAXIMUS XI Formula

Current temps and clocks are, 53/53/53/52/52/52/51/51 (in order by core) 1.45v (adaptive) with 98°-105° under stress testing and 86° under normal usage (video games, rendering, ect)
 
Last edited:
Well, for the i9-9900K, an AIO isn't going to do much good for overclocking.

My setup:
Processor: Intel Core i7-8700K
Clock Speed: 5100 MHz on 6 cores
Cooler: NZXT Kraken x62
Max Temperatures under load (Prime95 v26.6): 82C

Note: I use version 26.6 primarily because any newer version of Prime95 is unrealistically stressful for someone who just plays games (me).

Note 2: I have 3 ML120 Fans blowing air from outside the case (ambient temperature of 20C) directly on the radiator and 2 of the stock fans that help pull the air through.

Note 3: I replaced the thermal compound that came with the Kraken with some of my leftover noctua thermal paste from my NH-D15.

I don't see over 80C in most games. But if I am hitting 82C on a six core processor at 5100 MHz, and the i9-9900K comes stock with a turbo of 5 GHz, if you overclock it to 5.1 GHz you will probably want a custom water loop. Of course, my radiator isn't among the bigger options out there, but I would still recommend a custom water cooling loop if you intend on overclocking the i9-9900K.
 
Last edited:
If you want to overclock on a 9900k, it would be wise to go with a custom loop. If you HAVE to make due with an AIO, I would highly recommend that you not look at anything less than a 360mm unit and that you factor in a very good case along with plenty of case fans as the major problem may not be the CPU itself but the well known problem with VRM throttling due to the power delivery demands on that configuration and the fact that the second you switch to water you are removing any chance that the residual airflow from an air cooler is going to help with that by passing over the voltage regulation components. Making sure you have exceptional case airflow will help somewhat to mitigate that, but you still might encounter problems unless you go with a very high end board.

Overclocking on the 9900k is simply not very feasible as the stock operation of that chip is already pretty well maxed out as far as the thermal headroom is concerned.

Simply not worth the cost or effort for a 5-8% increase in performance.
 
Dec 12, 2018
26
0
30
Well, for the i9-9900K, an AIO isn't going to do much good for overclocking.

My setup:
Processor: Intel Core i7-8700K
Clock Speed: 5100 MHz on 6 cores
Cooler: NZXT Kraken x62
Max Temperatures under load (Cinebench R20): 82C

Note: I use Cinebench primarily because Prime95 is unrealistically stressful for someone who just plays games (me).

Note 2: I have 3 ML120 Fans blowing air from outside the case (ambient temperature of 20C) directly on the radiator and 2 of the stock fans that help pull the air through.

Note 3: I replaced the thermal compound that came with the Kraken with some of my leftover noctua thermal paste from my NH-D15.

I don't see over 80C in most games. But if I am hitting 82C on a six core processor at 5100 MHz, and the i9-9900K comes stock with a turbo of 5 GHz, if you overclock it to 5.1 GHz you will probably want a custom water loop. Of course, my radiator isn't among the bigger options out there, but I would still recommend a custom water cooling loop if you intend on overclocking the i9-9900K.


you understand that an 8700k statistically runs 1/4x hotter than a 9900k right?
 
Dec 12, 2018
26
0
30
If you want to overclock on a 9900k, it would be wise to go with a custom loop. If you HAVE to make due with an AIO, I would highly recommend that you not look at anything less than a 360mm unit and that you factor in a very good case along with plenty of case fans as the major problem may not be the CPU itself but the well known problem with VRM throttling due to the power delivery demands on that configuration and the fact that the second you switch to water you are removing any chance that the residual airflow from an air cooler is going to help with that by passing over the voltage regulation components. Making sure you have exceptional case airflow will help somewhat to mitigate that, but you still might encounter problems unless you go with a very high end board.

Overclocking on the 9900k is simply not very feasible as the stock operation of that chip is already pretty well maxed out as far as the thermal headroom is concerned.

Simply not worth the cost or effort for a 5-8% increase in performance.

right now i have a Kracken x72, im switching to an EVGA CLC 280 and right now with a 53/53/53/52/52/52/51/51 1.45v OC my CPU runs at around 98°-105° under stress test, and 86° under normal usage. Ive been told the EVGA CLC 280 has a much better system that compensates for not having a 360MM system. Ill be running it in Push pull config instead of pull config like i have now. You are correct though, i do see the VRMs get decently hot. I do plan to run a custom loop eventually. Check my OP i updated my current specs.
 
Well, you're about 18-25°C PAST the acceptable safe thermal envelope, so, that's a problem. There is no such "thing" as "compensating" for not having a 360mm system. There is nothing that could be done TO compensate for the fact that a 360mm system has a lot more radiator surface area, and more fan surface area, and a 280mm system does not. I don't know where you heard that at, but that's patently ludicrous, and I'd stop listening to or getting information from whoever told you, or wherever you read it at.

It might be worth looking into a home brewed modification allowing a directly blowing VRM fan to be implemented. There are actually coolers like some of the Cryorig AIO units that include a fan that can be pointed directly at the VRMs, but again those are probably not coolers with sufficient base cooling capacity for overclocking the 9900k.

As far as the 9900k supposedly running cooler due to the change in internal TIM to solder, yeah, it really hasn't panned out that way in the real world. In fact, a lot of indications are that due to the compromises that had to be made in order to implement that change, they actually might be worse in some scenarios. But regardless, the fact is they generally are ALREADY binned CPUs that are already at the end of their leash in terms of thermal and stability overhead, and I have yet to see anybody overclock one as a daily driver to any extent past the stock behavior that could be thermally managed well while also providing enough of an increase to even bother with.

At the stock configuration, they are ok. Past that, they are getting into FX-9590 territory with unmanageable thermals and VRM problems.
 
Dec 12, 2018
26
0
30
Well, you're about 18-25°C PAST the acceptable safe thermal envelope, so, that's a problem. There is no such "thing" as "compensating" for not having a 360mm system. There is nothing that could be done TO compensate for the fact that a 360mm system has a lot more radiator surface area, and more fan surface area, and a 280mm system does not. I don't know where you heard that at, but that's patently ludicrous, and I'd stop listening to or getting information from whoever told you, or wherever you read it at.

It might be worth looking into a home brewed modification allowing a directly blowing VRM fan to be implemented. There are actually coolers like some of the Cryorig AIO units that include a fan that can be pointed directly at the VRMs, but again those are probably not coolers with sufficient base cooling capacity for overclocking the 9900k.

As far as the 9900k supposedly running cooler due to the change in internal TIM to solder, yeah, it really hasn't panned out that way in the real world. In fact, a lot of indications are that due to the compromises that had to be made in order to implement that change, they actually might be worse in some scenarios. But regardless, the fact is they generally are ALREADY binned CPUs that are already at the end of their leash in terms of thermal and stability overhead, and I have yet to see anybody overclock one as a daily driver to any extent past the stock behavior that could be thermally managed well while also providing enough of an increase to even bother with.

At the stock configuration, they are ok. Past that, they are getting into FX-9590 territory with unmanageable thermals and VRM problems.

for benching yes im past the acceptable "safe" thermal envelope, but for realistic gameplay and rendering i am at a safe thermal level. Right now what i think is happening with my X72 is the IHS is not making good enough contact with the AIO plate, as a result the thermal transfer isnt as good as it should be. im testing the X72 and the CLC 280 and ill see what my temperatures are with either or.
 
for benching yes im past the acceptable "safe" thermal envelope, but for realistic gameplay and rendering i am at a safe thermal level. Right now what i think is happening with my X72 is the IHS is not making good enough contact with the AIO plate, as a result the thermal transfer isnt as good as it should be. im testing the X72 and the CLC 280 and ill see what my temperatures are with either or.


If you are past the recommend thermal envelope while running stress, then you are past the thermal envelope PERIOD. There is no "safe for realistic game play" vs "benching". There is ONLY "what is my maximum thermal response under full load conditions". "Gaming" or "applications" can in some cases load up the CPU just as much as any "bench" or "stress test". It is ALL just software, and what one software can do, another can do. What are you testing your thermal compliance with?

If you are using ANYTHING other than Prime95 version 26.6 to test your thermal compliance then you are 120% most probably not testing correctly or accurately. Other versions of Prime for example, use AVX instructions which are neither steady state nor realistic. Prime 26.6 does not use AVX instructions and is both steady state AND realistic.

Chances are good if you are using some other utility to test thermal compliance then you are getting a result that is either unrealistically high or in some cases, like with Aida, no where near TDP and therefore useless.



*Quick and dirty overview of overclocking/stability validation procedure.

Set CPU multiplier and voltage at desired settings in BIOS. Do not use presets or automatic utilities. These will overcompensate on core and other voltages. It is much better to configure most core settings manually, and leave anything left over on auto until a later point in time if wish to come back and tweak settings such as cache (Uncore) frequency, System agent voltage, VCCIO (Internal memory controller) and memory speeds or timings (RAM) AFTER the CPU overclock is fully stable.

Save bios settings (As a new BIOS profile if your bios supports multiple profiles) and exit bios.

Boot into the Windows desktop environment. Download and install Prime95 version 26.6.

Download and install either HWinfo or CoreTemp.

Open HWinfo and run "Sensors only" or open CoreTemp.

Run Prime95 (ONLY version 26.6) and choose the "Small FFT test option". Run this for 15 minutes while monitoring your core/package temperatures to verify that you do not exceed the thermal specifications of your CPU.

(This should be considered to be 80°C for most generations of Intel processor and for current Ryzen CPUs. For older AMD FX and Phenom series, you should use a thermal monitor that has options for "Distance to TJmax" and you want to NOT see distance to TJmax drop below 10°C distance to TJmax. Anything that is MORE than 10°C distance to TJmax is within the allowed thermal envelope.)

If your CPU passes the thermal compliance test, move on to stability.

Download and install Realbench. Run Realbench and choose the Stress test option. Choose a value from the available memory (RAM) options that is equal to approximately half of your installed memory capacity. If you have 16GB, choose 8GB. If you have 8GB, choose 4GB, etc. Click start and allow the stability test to run for 8 hours. Do not plan to use the system for ANYTHING else while it is running. It will run realistic AVX and handbrake workloads and if it passes 8 hours of testing it is probably about as stable as you can reasonably expect.

If you wish to check stability further you can run 12-24 hours of Prime95 Blend mode or Small FFT.

You do not need to simultaneously run HWinfo or CoreTemp while running Realbench as you should have already performed the thermal compliance test PLUS Realbench will show current CPU temperatures while it is running.

If you run the additional stability test using Prime95 Blend/Small FFT modes for 12-24 hours, you will WANT to also run HWinfo alongside it. Monitor HWinfo periodically to verify that no cores/threads are showing less than 100% usage. If it is, then that worker has errored out and the test should be stopped.

If you find there are errors on ANY of the stability tests including Realbench or Prime95, or any other stress testing utility, you need to make a change in the bios. This could be either dropping the multiplier to a lower factor or increasing the voltage while leaving the multiplier the same. If you change voltage or multiplier at ANY time, you need to start over again at the beginning and verify thermal compliance again.



A more in depth but general guide that is still intended for beginners or those who have had a small amount of experience overclocking can be found here:

*CPU overclocking guide for beginners
 
Dec 12, 2018
26
0
30
If you are past the recommend thermal envelope while running stress, then you are past the thermal envelope PERIOD. There is no "safe for realistic game play" vs "benching". There is ONLY "what is my maximum thermal response under full load conditions". "Gaming" or "applications" can in some cases load up the CPU just as much as any "bench" or "stress test". It is ALL just software, and what one software can do, another can do. What are you testing your thermal compliance with?

If you are using ANYTHING other than Prime95 version 26.6 to test your thermal compliance then you are 120% most probably not testing correctly or accurately. Other versions of Prime for example, use AVX instructions which are neither steady state nor realistic. Prime 26.6 does not use AVX instructions and is both steady state AND realistic.

Chances are good if you are using some other utility to test thermal compliance then you are getting a result that is either unrealistically high or in some cases, like with Aida, no where near TDP and therefore useless.



*Quick and dirty overview of overclocking/stability validation procedure.

Set CPU multiplier and voltage at desired settings in BIOS. Do not use presets or automatic utilities. These will overcompensate on core and other voltages. It is much better to configure most core settings manually, and leave anything left over on auto until a later point in time if wish to come back and tweak settings such as cache (Uncore) frequency, System agent voltage, VCCIO (Internal memory controller) and memory speeds or timings (RAM) AFTER the CPU overclock is fully stable.

Save bios settings (As a new BIOS profile if your bios supports multiple profiles) and exit bios.

Boot into the Windows desktop environment. Download and install Prime95 version 26.6.

Download and install either HWinfo or CoreTemp.

Open HWinfo and run "Sensors only" or open CoreTemp.

Run Prime95 (ONLY version 26.6) and choose the "Small FFT test option". Run this for 15 minutes while monitoring your core/package temperatures to verify that you do not exceed the thermal specifications of your CPU.

(This should be considered to be 80°C for most generations of Intel processor and for current Ryzen CPUs. For older AMD FX and Phenom series, you should use a thermal monitor that has options for "Distance to TJmax" and you want to NOT see distance to TJmax drop below 10°C distance to TJmax. Anything that is MORE than 10°C distance to TJmax is within the allowed thermal envelope.)

If your CPU passes the thermal compliance test, move on to stability.

Download and install Realbench. Run Realbench and choose the Stress test option. Choose a value from the available memory (RAM) options that is equal to approximately half of your installed memory capacity. If you have 16GB, choose 8GB. If you have 8GB, choose 4GB, etc. Click start and allow the stability test to run for 8 hours. Do not plan to use the system for ANYTHING else while it is running. It will run realistic AVX and handbrake workloads and if it passes 8 hours of testing it is probably about as stable as you can reasonably expect.

If you wish to check stability further you can run 12-24 hours of Prime95 Blend mode or Small FFT.

You do not need to simultaneously run HWinfo or CoreTemp while running Realbench as you should have already performed the thermal compliance test PLUS Realbench will show current CPU temperatures while it is running.

If you run the additional stability test using Prime95 Blend/Small FFT modes for 12-24 hours, you will WANT to also run HWinfo alongside it. Monitor HWinfo periodically to verify that no cores/threads are showing less than 100% usage. If it is, then that worker has errored out and the test should be stopped.

If you find there are errors on ANY of the stability tests including Realbench or Prime95, or any other stress testing utility, you need to make a change in the bios. This could be either dropping the multiplier to a lower factor or increasing the voltage while leaving the multiplier the same. If you change voltage or multiplier at ANY time, you need to start over again at the beginning and verify thermal compliance again.



A more in depth but general guide that is still intended for beginners or those who have had a small amount of experience overclocking can be found here:

*CPU overclocking guide for beginners


After swapping to a CLC 280, i saw a drop in 10°. New metrics are 89°-94° under load with cinebench and while using Prime95 i get 91°-96°. While doing a CPU intensive render the highest i saw my CPU go was 81° now. While gaming i don't think ive seen it go above 70° so i think thats a huge improvment from the Krackenx72
 

Notorious^

Reputable
Feb 17, 2019
104
18
4,615
Asking what is the best [ENTER PC PART/BRAND HERE] is like asking whats the best motor oil to use in a car. Everyone has a opinion and everyone's opinion is the "best".

I will simply tell you i overclock to 5.1 GHz on a 280mm Corsair Liquid Cooler with no issues. Idle is around 34-35c on 5.1GHz and full load i see around 75-78c. I do have a custom fan curve profile setup on my cooler as well as i removed the pre applied paste and swapped it for Grizzly. I just game though, nothing else. Like you, i researched for weeks before buying parts to go with my 9900K. i watched hours of YouTube videos, read countless articles and discovered that people are cooling 9900K's just fine with air and liquid coolers, granted the higher end coolers.

You won't need anything bigger than a 280mm unless you're doing something massive on a crazy overclock.
 
Yeah, that's 150% dubious, at best. ANY overclocking on a 9900k should have either one of maybe five top shelf air coolers or at minimum an extra large 280mm AIO with a thicker radiator for more surface area. A 320mm would be better. A custom loop would be even better.

If you are sitting on a full time, all core 5.1Ghz overclock on that 9900k, and you are only hitting 78°C, then you are either living in a cave or refrigerated room that has an insanely low ambient, or you are unstable, because there is no realistic way you could possibly be getting that low of temps with a plain Jane 280mm cooler otherwise, unless your voltage is so low that the stability is suffering.

Especially if you're running high speed memory. I'd almost be willing to bet money that system cannot make it through an 8 hour stress test run of Realbench, even with only half your memory capacity selected, without throwing errors. Stability doesn't mean "not getting blue screens or freezes", it means, not getting errors, AT ALL. And if you can't do that, then you need to increase the voltage, in which case, you are not going to have those temps. I'm calling BS, sorry.

I've seen a lot of people, many samples, trying to maintain a 5Ghz all core OC on the 9900k, and some can do it, many in fact, but they either have temps above what we recommend as safe for a daily driver or they have advanced cooling configurations. I guess it's possible you have the utmost golden sample CPU ever to leave the foundry, but I think it's doubtful.
 

Notorious^

Reputable
Feb 17, 2019
104
18
4,615
Well that was a exceptional report my friend. That read like you're trying to convince yourself that all that is 100% defacto gods word.

When you're ready you can book an airline flight on out to me, i'll put you up in the guest house for the weekend and show you a few things.

Indeed my rig is stable 5.0, 5.1 GHz, sorry to have triggered your feelings. I was going to make a video for my channel and dedicate it to you with testing results but i kinda decided to go back on living my life.

I was trying to help the OP not get frustrated with choosing a cooler to keep the 9900K cool for his needs as we all started out newbies and was trying to help a brother out. But heaven forbid someone who has tested, tested and tested some more say something that you don't agree with lol.

I will hand it back over to you boss. Get him all fixed up, ok.

Cheers ././\.\.
 
Dec 12, 2018
26
0
30
Asking what is the best [ENTER PC PART/BRAND HERE] is like asking whats the best motor oil to use in a car. Everyone has a opinion and everyone's opinion is the "best".

I will simply tell you i overclock to 5.1 GHz on a 280mm Corsair Liquid Cooler with no issues. Idle is around 34-35c on 5.1GHz and full load i see around 75-78c. I do have a custom fan curve profile setup on my cooler as well as i removed the pre applied paste and swapped it for Grizzly. I just game though, nothing else. Like you, i researched for weeks before buying parts to go with my 9900K. i watched hours of YouTube videos, read countless articles and discovered that people are cooling 9900K's just fine with air and liquid coolers, granted the higher end coolers.

You won't need anything bigger than a 280mm unless you're doing something massive on a crazy overclock.


Yah I realize that, I was asking to see if anyone knows of a post or guide that has tested, thoroughly, AIOs and Custom loops against the 9900K. All I've seen are personal builds and what they have.

I also have a custom fan profile. 100% pump speed at all times, 25-100% fan speed between 30°-60°c
 
The funny thing is, he already CHOSE a cooler, before you ever posted anything. Which means any advice you, or I, might have offered on the subject was not going to be a factor regardless. And you don't need to talk down to me like you are somebody's parent. I'd be really surprised if you were even half my age, and I've been working with just about every kind of computer system that exists since before most our members were born. Now, that doesn't necessarily mean that I know anything, and compared to many of our other moderators and some members, I don't know squat, but I assure you that it is very, VERY unlikely were I to come to you house there would be much you could teach me. Maybe, but probably not.

Aside from all that, my comments were not intended to be a direct insult to you or your system, only that many people labor under the illusion that their systems are somehow capable of something that everybody else's is not, until they realize that the methodology they are using to determine and validate the thermal compliance and stability is hopelessly flawed and not in line with the reality of the process. Once the proper methods are used they usually tend to discover that what they though WAS, is not, and that just like everybody else there are physical laws that govern how well a thing can do given the restrictions of the silicon and hardware, and cooling. But if you believe you are good, that is your prerogative, and has no bearing on the advice I will give to others based on best practices AND the general results being seen by everybody else with similar configurations. Maybe you got the Willy Wonka golden sample ticket, but as I said, I'm still dubious. I have no doubt that you can run your system at that frequency and achieve those thermal results, I just don't believe that your system is fully stable while doing so. You run the two following tests for 8 hours each and if there are zero errors then I guess I'll have to agree that you have an extremely terrifically binned chip.

Realbench stress test option, full memory capacity, 8 hours.

Prime custom stress test, as follows.

Final testing with Prime95

It is highly advisable that you do a final test using Prime95 version 26.6 (And ONLY version 26.6 except as noted below) choosing the Custom test. You can also use the Blend mode option but after a fair amount of personal testing, asking questions from some long time members with engineering level degrees that have forgotten more about memory architectures than you or I will ever know, and gathering opinions from a wide array of memory enthusiasts around the web, I'm pretty confident that the custom option is a lot more likely to find errors with the CPU and memory configuration, and faster, if there are any to be found.

Please note as this is rather important, if you prefer, or have problems running version 26.6 because you have a newer platform that doesn't want to play nice with version 26.6, you can use the latest version of Prime95 with the Custom test selected but you will need to make the following change.

If you wish to use a newer version than 26.6 make the following edit to the "local.txt" file located in the Prime95 folder.

Find the line value that specifies CpuSupportsAVX=1, and change it to CpuSupportsAVX=0

Then click File-->Save, and then close the document.

Now open Prime95.

Click on "Custom". Input a value of 512k in the minimum FFT size field. Leave the maximum FFT size field at 4096k. In the "Memory to use" field you should take a look at your current memory allocation in either HWinfo or system resource monitor. Whatever "free" memory is available, input approximately 75% of that amount. So if you currently have 16GB of installed memory, and approximately 3GB are in use or reserved leaving somewhere in the neighborhood of 13GB free, then enter something close to 75% of that amount.

So if you have 13GB free, or something reasonably close to that, then 75% of THAT would be 9.75GB, which, when multiplies times 1024 will roughly equal about 9984MB. You can average things out by simply selecting the closest multiple of 1024 to that amount just to keep it simple, so we'll say 10 x 1024= 10240mb and enter that amount in the field for "Memory to use (MB)". We are still well within the 13GB of unused memory BUT we have left enough memory unused so that if Windows decides to load some other process or background program, or an already loaded one suddenly needs more, we won't run into a situation where the system errors out due to lack of memory because we've dedicated it all to testing.

I've experienced false errors and system freezes during this test from over allocating memory, so stick to the method above and you should be ok.


Moving right along, do not change the time to run each FFT size.Leave that set to 15 minutes.

Click run and run the Custom test for 8 hours. If it passed Memtest86 and it passes 8 hours of the Custom test, the memory is 100% stable, or as close to it as you are ever likely to get but a lot of experts in the area of memory configuration suggest that running the extended Windows memory diagnostic test is also a pretty good idea too.

If you get errors, (and you will want to run HWinfo alongside Prime95 so you can periodically monitor each thread as Prime will not stop running just because one worker drops out, so you need to watch HWinfo to see if there are any threads not showing 100% usage which means one of the workers errored and was dropped) then you need to either change the timings, change the DRAM voltage or change the DRAM termination voltage, which should be approximately half of the full DRAM voltage.

There are also other bios settings that can affect the memory configuration AND stability, such as the VCCIO and system agent voltages, so if you have problems with stability at higher clock speeds you might want to look at increasing those slightly. Usually, for Intel at least, something in the neighborhood of 1.1v on both those is pretty safe. There are a substantial number of guides out there covering those two settings, but most of them are found within CPU overclocking guides so look there in guides relevant to your platform.

As a further measure of assurance that your WHOLE configuration is stable, you can download and run Realbench for 8 hours. If the system freezes or fails when running Realbench with your full memory amount set, try running it again but select only half your amount of installed memory.
 
Dec 12, 2018
26
0
30
Actually, I have quite a stock of parts. So that's not true, If someone had posted some really good results on a specific cooler I'd of bought it and tried it out as well.
 
Okay, before we get into hurt feelings about stability and who's word is correct, remember we are all here to assist someone - without compensation or economic return.

Now, Darkbreeze is correct. Prime95 V26.6 (Doesn't use AVX instructions which cause a ton of heat) is a good test to run while you are sleeping or away from your computer for 6-8 hours. The reason it is recommended to run these stability tests for this long is to prevent failure in your machine. Imagine being in the middle of an 8 hour gaming session and you go to save your progress, the system crashes and you lose all of that data. 100 MHz on an overclock is only really worth bragging rights, and doesn't make a real difference in a game, unless it is an older CPU with low core and thread count, which a 9900K is clearly above and beyond. Now if you use AVX instructions, a newer version of Prime95 is entirely viable.

Also, 80C is the maximum you want to see on any device that is overclocked or under load. The reason is above this thermal limit, you will see a decrease in longevity in your processor. So if you are overclocking from 5 GHz to 5.1 GHz from a safe range of 70C - 80C on an AIO or air cooler to an unsafe range above 80C, you either need to increase your cooling capacity or dial it back. The general rule of thumb is the more you increase the clock speed, the requirements in voltage will increase beyond the value of the speed increase.
A good example:
1.20 @ 4.9 GHz
1.30 @ 5.0 GHz
1.50 @ 5.1 GHz

The best thing to do would to be to find the best range at around 1.30 - 1.40, which is usually the highest you can go before cooling becomes extremely difficult.

This is why both Darkbreeze and I have both recommended the OP to use a custom water cooling loop for best thermal and performance results, as not only will you be able to push your machine higher than before, but you will also have better have temperatures.

Now I am not saying that we know everything, but remember we are here to help.

I have built dozens of computers from the 2005 era to a brand new computer in 2018. I started PC building in 2018, and yes, I have much to learn still. Darkbreeze is among one of the several people on this site who have helped me along the way.
 
Dec 12, 2018
26
0
30
Okay, before we get into hurt feelings about stability and who's word is correct, remember we are all here to assist someone - without compensation or economic return.

Now, Darkbreeze is correct. Prime95 V26.6 (Doesn't use AVX instructions which cause a ton of heat) is a good test to run while you are sleeping or away from your computer for 6-8 hours. The reason it is recommended to run these stability tests for this long is to prevent failure in your machine. Imagine being in the middle of an 8 hour gaming session and you go to save your progress, the system crashes and you lose all of that data. 100 MHz on an overclock is only really worth bragging rights, and doesn't make a real difference in a game, unless it is an older CPU with low core and thread count, which a 9900K is clearly above and beyond. Now if you use AVX instructions, a newer version of Prime95 is entirely viable.

Also, 80C is the maximum you want to see on any device that is overclocked or under load. The reason is above this thermal limit, you will see a decrease in longevity in your processor. So if you are overclocking from 5 GHz to 5.1 GHz from a safe range of 70C - 80C on an AIO or air cooler to an unsafe range above 80C, you either need to increase your cooling capacity or dial it back. The general rule of thumb is the more you increase the clock speed, the requirements in voltage will increase beyond the value of the speed increase.
A good example:
1.20 @ 4.9 GHz
1.30 @ 5.0 GHz
1.50 @ 5.1 GHz

The best thing to do would to be to find the best range at around 1.30 - 1.40, which is usually the highest you can go before cooling becomes extremely difficult.

This is why both Darkbreeze and I have both recommended the OP to use a custom water cooling loop for best thermal and performance results, as not only will you be able to push your machine higher than before, but you will also have better have temperatures.


I appreciate your input, im aware of the longevity reduction but their has been THOROUGH testing done on longevity and testing of a CPU. a CPU running at 90° for 8 hours a day every day for an entire year still lasted 5-8 years and so no HUGE loss in real longevity. and seeing as i build my computers to last me around 5 years before i upgrade them than seeing the OCCASIONAL 90-105° when benching and testing ONLY i see no real harm in running my current overclock that never gets above 70-80° (SO FAR) i see no real downfall from an occasional spike. What i look for in stability tests is not temperature but actual stability. I ran Prime95 V26.6 last night for aprox. 8.5 hours and no issues, and now that i found a comfortable overclock (the one stated in my OP) i have no need to push the envelope anymore, however, if switch to a custom loop i will TRY to get a stable 53 all core overclock that uses both safe(ish) voltages and running temps.

id like to note that the prime95 i ran with the KrackenX72 for 8 hours had a peak temp of 108° with an average of 104°. Last nights Prime95 run with the EVGA CLC 280 ran with a peak of 102° with an avg of 96°
 
I appreciate your input, im aware of the longevity reduction but their has been THOROUGH testing done on longevity and testing of a CPU. a CPU running at 90° for 8 hours a day every day for an entire year still lasted 5-8 years and so no HUGE loss in real longevity. and seeing as i build my computers to last me around 5 years before i upgrade them than seeing the OCCASIONAL 90-105° when benching and testing ONLY i see no real harm in running my current overclock that never gets above 70-80° (SO FAR) i see no real downfall from an occasional spike. What i look for in stability tests is not temperature but actual stability. I ran Prime95 V26.6 last night for aprox. 8.5 hours and no issues, and now that i found a comfortable overclock (the one stated in my OP) i have no need to push the envelope anymore, however, if switch to a custom loop i will TRY to get a stable 53 all core overclock that uses both safe(ish) voltages and running temps.

id like to note that the prime95 i ran with the KrackenX72 for 8 hours had a peak temp of 108° with an average of 104°. Last nights Prime95 run with the EVGA CLC 280 ran with a peak of 102° with an avg of 96°

Okay, but you realize this can cause damage to the motherboard and memory as well? If so - then I suppose that is your choice. Just remember water cooling the machine is cheaper than replacing the entire machine.
 
I appreciate your input, im aware of the longevity reduction but their has been THOROUGH testing done on longevity and testing of a CPU. a CPU running at 90° for 8 hours a day every day for an entire year still lasted 5-8 years and so no HUGE loss in real longevity. and seeing as i build my computers to last me around 5 years before i upgrade them than seeing the OCCASIONAL 90-105° when benching and testing ONLY i see no real harm in running my current overclock that never gets above 70-80° (SO FAR) i see no real downfall from an occasional spike. What i look for in stability tests is not temperature but actual stability. I ran Prime95 V26.6 last night for aprox. 8.5 hours and no issues, and now that i found a comfortable overclock (the one stated in my OP) i have no need to push the envelope anymore, however, if switch to a custom loop i will TRY to get a stable 53 all core overclock that uses both safe(ish) voltages and running temps.

id like to note that the prime95 i ran with the KrackenX72 for 8 hours had a peak temp of 108° with an average of 104°. Last nights Prime95 run with the EVGA CLC 280 ran with a peak of 102° with an avg of 96°


Just because "A" CPU did that, IF it did, and so far we just have your word that such a test was even performed (In fact, could NOT have been performed, since there weren't any 9900k's 5-8 years ago), but even assuming a different CPU model (Which cannot be assumed to have the same expectation or longevity since the implementation of the architecture is aggressively different especially in the area of the heat spreader, thermal interface and CPU die thickness when looking at the Coffee lake refresh models versus ALL prior Intel models including Ivy bridge and older units that had soldered TIM) has a similar result, it is still only ONE sample.

There are people who have dramatically shorted their hardware and got away with it. Doesn't mean it's going to happen for most people though.

If you want to read the conclusions of a LONG term project, then I'd advise you to read this, in it's ENTIRETY.



I'd also like to see the 8 year 90°C project and it's conclusions, if there is one and you can provide a link to it. I don't believe I've ever seen or even heard hints of any such project, but I reserve judgement until having read it if you can provide the link.
 
Last edited:
Mar 31, 2019
55
5
35
I deleted my post because it could cause arguments

My opinion is that if you have got such expensive 9900k cpu you could spend another 200 $ or so to get a custom water cool setup i know i have seen people on internet who have successfully oc the 9900k with custom loop and stability is also allrite
 
Jan 20, 2020
4
0
10
So, I have a i9-9900k running in an Asus Maximus xi hero with a 7 year old Corsair H100 v1 with push/pull noctua's. My i9 is locked at a multiplier of x52 (5.2GHz) on ALL cores and I never hit temps above 70°c. So, I don't know what everyone here's theorizing about. You don't need a open loop or unicorn p!$$ to run normally, just a sense of knowledge in how to properly OC.
 
You're not talking to some twelve year olds on Reddit my friend. I've been overclocking everything that could be overclocked for the last, I dunno, fifteen to twenty years. I assure you that I DO have plenty of overclocking knowledge. Am I at a competitive level? No, absolutely not. But I'm certainly no greenhorn either. Admittedly, I've not worked with a 9900k myself, so I do not have any personal hands on experience in that regard, but when practically every accomplished and competitive overclocker out there recommends open loop cooling if you're going to try to overclock, along with as high end of a motherboard as you can afford, since that's not something they've normally been known to throw out there, you kind of want to believe THEM over you.
 

Phaaze88

Titan
Ambassador
So, I have a i9-9900k running in an Asus Maximus xi hero with a 7 year old Corsair H100 v1 with push/pull noctua's. My i9 is locked at a multiplier of x52 (5.2GHz) on ALL cores and I never hit temps above 70°c. So, I don't know what everyone here's theorizing about. You don't need a open loop or unicorn p!$$ to run normally, just a sense of knowledge in how to properly OC.
I have to question what you did to confirm:
-voltage stability
-thermal stability
 
  • Like
Reactions: Darkbreeze