XPS 8960 CPU & Cooler Upgrade: i7-13700 to i9-14900K, Dell Advanced Tower & Dell Liquid Cooler.

With the i9 14900K being installed,  finished all the the tests on both coolers and wanted to test the coolers with overclocking.

I can go to bios and enable overclocking with the K CPU, I select the one and only level wich is  OC level 1, but there is no performance gain at all. 

I see the CPU temperature is allowed to go maybe 10-15% higher but at the end, none  of the 5 benchmarks I used shows any performance gain from this option at all.

I can`t find any topic explaining this in detail how this option works exactly.

FWIW: I have the 8960 with i7-13700K and the tower cooler - what Dell calls "Advanced cooling".

"I really wanted to know what`s the difference between all three Dell cooler options regards of performance and noise."

One thing everyone will agree on is to avoid the standard 'pancake' cooler with an i7 or i9.

I did a lot of research on Dell's water cooler vs. tower cooler, both before and after I bought my PC. There's no obvious answer as to which is better, it depends on your use case.

In my use cases, I am very unlikely to push my processor close to its full capabilities. So for me, the question was: Which one is quieter? I read widely, both here - from posters I trust - and outside sources, and the consensus was that the tower cooler is more quiet. That's why I bought the 'K' type, because it came with the tower cooler, even though I'm not one to overclock.

On the other hand, nobody said that the water cooler was noisy, only that it was audible as a background hum or gurgle. You might not even notice it, but for me, I'm happy with my choice.

As water coolers go, you probably know that Dell's water cooler is a rather low rent version. It only has one fan, and the radiator is only 120mm. Any third-party water cooler, even an all-in-one model, is going to have at least 2 fans with 240mm radiators. So while Dell's water cooler is adequate, it's not in the same league as a third-party model.

Thanks to both of you! Turning on the Dell Bios overclocking LV1 feature is resulting in a minimal but measurable performance drop. Using the suggested Intel Extreme Tuning Utility auto overclock has pretty much no effect, however using it manually was able to get a solid 15-20% performance increase.

 This requires way more testing, learning, and understating …  I will not include overclocking in the last video for sure :)

@ProfessorW00d​

Before I saw your reply I realized that there is space here for something  better.

But than again , I don’t want to looking for bolts,  spacers washers  and  workarounds because Dell made the cooler mounting points none standard. This pc is owned by a client so I have to stick with Dell parts. 
Thanks! 

If you want to see the difference between them run a prime 95 torture test.  the pancake and the Dell tower cooler will just throttle, the pancake cooler isn't even the same size as the intel stock pancake cooler it's smaller, tbh they just need to get rid of the pancake one and offer either tower or water. 

Finally finished all the tests with the Standard & Advanced Air Coolers and the 125W Liquid cooler.  With the i9-14900K same result as the i7-13700. Performance wise, there is no difference between the tower cooler and the pump.

However something happened to the pump, it started smelling … Ended up making a 30 minute video, included all the things  I ran into with this upgrade.

https://youtu.be/HFBwlwSPwy0?si=pQi5aYvh-QNrN-Fy

@S.Tech9​ This will sound harsh, but your testing method is 100% wrong. You don't use Cinebench numbers to say this cooler is better. That is not even telling you the story. Unless you can overclock the CPU, regardless of the cooler you are going to be within 3% anyway, You need to do a load test and look at the actual THERMALS put out by the CPU, between the pancake cooler, tower and the water cooler. 

You say there is 0 difference, well that is because your testing method doesn't measure the cooler preference. 

Here is an example, I ran the same bench you did.  DELL Pancake Cooler, Dell  Tower Cooler and Nactua D12L. 

The pancake cooler hit Max TDP at the start of the test and thermal throttled down to 2.6 GHz For the remainder of the test, test results for Multicore were 10400, with an average temperature of 95c 

Dell Tower Cooler hit a max  TPD of 98c With an average of 90c The average speed was 3.3 GHz throughout the test with a score of  10800

The Noctura Cooler, which faired the best, hit a maximum TDP of 81C with an average CPU Temperature of 70C, the core frequency averaged 4.1ghz, and a test score of 11600.

While the test scores were within 3 %, the Effective cooling was a 5c Delta between the Pancake and Dell Tower Coolers and a 15 - 20c Delta between the Pancake and Tower Coolers. While the actual benchmarks were about the same.  The cooling efficiency between the different coolers leaves a lot to be desired by the Dell coolers. While an aftermarket cooler may not give you extremely different benchmarks.  The heat dissipation over time and average lower temperatures will make sure your system does not throttle and keeps it performing for a long time to come. 

@JamieLinux​ 

"This will sound harsh, but your testing method is 100% wrong. " No worries, let`s figure out where I was wrong.

Starting from here, I am getting confused: 

"That is not even telling you the story. Unless you can overclock the CPU"

If you watched the whole video, then you know that Dell bios overclocking lv1 option & the Intel extreme tuning utility auto overclock feature did absolutely nothing on the i9-14900K CPU. With the manual overclocking option, I was able to gain some performance, but the fan was so loud, I ended up shutting it down. With these Dell cooler options; overclocking is just not happening with this XPS8960.

What is your overclocking Vs no overclocking score up with the two other coolers against the pancake ?

“The pancake cooler hit Max TDP at the start of the test and thermal throttled down to 2.6 GHz For the remainder of the test, test results for Multicore were 10400, with an average temperature of 95c 

Dell Tower Cooler hit a max  TPD of 98c With an average of 90c The average speed was 3.3 GHz throughout the test with a score of  10800

The Noctura Cooler, which faired the best, hit a maximum TDP of 81C with an average CPU Temperature of 70C, the core frequency averaged 4.1ghz, and a test score of 11600.”

Your numbers:

Pancake cooler      10400-> Dell Tower Cooler 10800 -> +3.8%
Pancake cooler      10400 -> Noctua D12L      11600      -> +11%

Dell tower cooler gave you +3.8% while the Noctua gave you +11%.
What CPU do you have and which version of Cinebench are you referring to?

i9-14900k, Cinebench R24,  Multicore test:

Pancake cooler 1452 -> Dell Tower Cooler 1654 -> +13%

Based on my numbers the Dell tower cooler gave me an extra 13% where you got 3.8%.
I wonder how the Noctua would perform in my case where you got +11% since as you are pointing it out” lower temperatures will make sure your system does not throttle” especially if you got 15-20c headroom from that cooler.

After running 250+ rounds of benchmarks I feel positive that I was able to measure the difference between the pancake, tower, and the liquid cooler accurately. If you still feel like I was 100% wrong, just let me know your exact method, see if I can replicate  it . 

@S.Tech9​ You are missing the point it was an example test, meaning I just wrote it out. You are not even giving the actual thermal values and using the bench numbers as the performance of the cooler, you can overclock the CPU with that system still 100 % does not negate the fact if you are testing coolers you need to give the thermal loads, not just the benchmark numbers and then claim there is 0 difference.  I can tell you with 100% fat, you release the thermal numbers the Dell Pancake and Tower cooler thermals are going to be a joke compared to the watercooler, Here I'll post a prime 95 torture test on a DELL XPS 8940 with a Nactua D12L and Furmark. Inside of DELL System

There is a Prime 95 Torture Test Maximum Temp on the package is 89c and Maximum on the core is 88 C with an Average Temperature of 68C.  That is in a room whose ambient temperature is 77 Degrees. 

If you tried that same test on the Dell Pancake Cooler and Tower cooler, it would 100% hit 100c and remain in the high 90s throughout the test, whereas the Dell Watercooler would not. 

Sorry my example was written in a way that made you think it was an actual test,  Here is an actual test. 

All that is in this case.

Anyway good luck on your YouTube,  my only suggestion is to do the numbers that include the thermal values, if you can't do that I don't know what to tell you. I would not pass off that the Dell Pancake and or Tower cooler are the same as the Water Cooler when they are 100 different. 

When testing coolers, you test the thermal load, and performance.

Lastly here is that same system with a 10 min torture test for Cinabench. 

10 mins with a score of13289 Maximum CPU Package and Core Temperatures 73C , Maximum Power Draw 162 Watts

The average temperature after the fans settled was 54C, Yes it was pumping hot air out the back.

What does this mean?  The Nactua Cooler keeps the CPU cooler under heavy load, while the system fans can replace 
all the hot air in the system, keeping it about 127 degrees or 59 degrees above room ambient temperature, while that may seem bad. The room doesn't have a fan or AC circulation which could have lowered those numbers. 

I tried to re-run this same test with the Dell pancake and tower cooler and got a DNF, as the thermals went so high that the system shut off. 

Test System Dell XPS 8940

                    I7 11700 

                    64 GB DDR 4000 MHZ memory running at 2933 Due to limitation of the motherboard.
                    Noctua D12l With 2 120 mm fans Noctua

                    92MM Scythe Exhaust Fan. 

                    120 MM SCythe HighStatic Pressure intake fan

                    80mm Noctua intake fan

@S.Tech9​ 

You can get Grub screws, 3x25mm (8)/ 3mm nylon locknuts (8)
Item #: 3962

Traxxas 3x25mm Grub Set Screws & 3mm Locknuts (rcsuperstore.com)

Install them into the case where the intel pancake cooler is installed, and then use any Noctua or any cooler that uses a standard intel mounting plate, you can just skip the mounting plate part for the LGA 1200,1700 and so forth and just use the included parts. 

Hi @JamieLinux 

would this be a reasonable mod for a beginner? I bought this system and it’s coming soon, would like to address these heat issues to make this system last. If possible, could you extend your guide with verbose additional details?

@Kfkfkfkfkfkfkfkdk​ Which system did you buy, just for clarity.