Please support our reinvestment into GamersNexus and its testing quality and accuracy via Patreon (www.patreon.com/gamersnexus ) or the GN Store (store.gamersnexus.net ) Article accompaniment (the video and article both have some unique portions): www.gamersnexus.net/guides/3561-cpu-cooler-testing-methodology-most-tests-are-flawed
With the emphasis on using the exact same hardware on every test and the demounting and remounting of the hardware so many times what are your concerns about breaking pins on the CPU?
“Guy who performs extreme RAM overclocks states thermal testing isn’t worth it”, thus, the reason for which the cooler testing scene is comparatively poor
I just want to say that I personally appreciate immensely the lengths to which you guys go to produce actual, reproducible and accurate data. In the last month I've looked at 6 reviews of a cooler, 1 of which put it ahead of the D15, 1 of which put it behind the D15, and 4 of which put it at about equal to the D15. This is VERY CLEARLY NOT A SIMPLE THING TO TEST FOR, and I applaud you for taking the Noctua approach. By which I mean: Noctua hasn't put out a new top of the line cooler since 2014, because they didn't feel like they could scientifically justify that they'd come up with an ACTUAL upgrade. The marketing department could have EASILY come up with at least 4 new coolers in the meantime, much like how many other manufacturers have. Now whether you actually buy Noctua's claims or not, the fact that they haven't introduced a new top of the line cooler since 2014 tells you something about how much they believe in their OWN product. (How many other companies outright tell you that they've worked on a fan design for 4 years, and are going to need at least 2 years more to make it perfect?) And you originally said you'd come back to cooler reviews years ago, yet didn't actually feel comfortable enough to do it until now, years later, tells me that you're not just trying to pull in views. This to me says that you want to actually do things right, rather than just push out numbers to boost your views. And for that I thank you.
1000000% this. THIS is why I trust GamersNexus above all else when it comes to hardware... Making actually reliable, reproducible data for something is NOT easy... and data quality is just not a thing you can hand wave to create. It takes a ton of know-how to do it right.
@@Mr.Morden well that is accounted with the test where GN does the cooler mounting pressure and contact . And if the cooler is mounted on Intel or and I understand the die size and location are different interms of monolithic and chiplet design but the coolers are not made Intel or AMD specific they are compatible
A large part of my job is testing equipment for the military. I feel like I just had my job explained me by someone who doesn't work there, and still understands it way better than me. Outstanding explanation.
Oddly if you want to find better methods then talk to someone who is in the fields of medical or food science based lab work studying organisms like listeria monocytogenes. They have to control temperatures quite accurately.
I work in research, this is why it is so necessary to know the test setup when evaluating results. Anytime anyone presents me with "new research indicates x" it always brings skepticism on my part. Good job Gamers Nexus!
These guys should be contracted. I mean my god. I have been a viewer for a long time. Sure they joke but no one else on the planet goes into this lvl of depth on a consistent basis.
rtings.com does amazing testing as well. Their TV testing is miles beyond anyone else. Unfortunately not a lot of PC stuff yet (they are just starting on keyboards and mice). It would be awesome if Gamers Nexus could create a similar exhaustive database of cooler and case reviews (and more, certainly), I'm sure everyone in the industry will send them their stuff.
Physics, real engineering, medical lab work, food science research into micro organisms. There's lots of professionalism out there, it just seems to end with PC manufacturers.
Gamers Nexus year 2050: "We now have complete control of weather globably so external temps should be more steady and while not perfect the data should be a bit more accurate"
I have a feeling that GN will be the company that most other companies and channels will start to go to for the true stats of products moving forward. It could even be a situation when people do reviews that they use a term like "Gamers Nexus Testings has found ...... " to report true validity in the numbers they cite. Super in depth and obviously alot of work goes into it.
@@JamesErath You don't see companies like Nvidia, Intel, AMD, TheVerge, LMG, etc saying "Go to buildzoid and/or GN", or "GN found this:", which is what hes saying. Yes WE say go to GN or buildzoid, but companies themselves do not
1:37 As a 25-year QA professional, this statement is accurate just about everywhere. Especially in software. Trust, but verify. Log everything. And remember: the translation programs work *FOR* the Matrix. ;)
I've done software QA for 1 year, and yeah. Everything behaves differently all the time, on every machine, and the same machine. God it could annoy the devs :D
Pretty damn superb. You might just be the first creator to win patreon dollars from me. Not for my own benefit. I kind of own the stuff I need to own and won't be buying for a few years. But this content is invaluable to new members of the community, and that I support wholly. Just never buy a teleprompter K. Ad hoc/page shuffling Steve is best Steve.
Steve and crew were the first Patreon support I ever offered to creators. Been a proud Patreon for more than 2 years now, and it's exactly because of content like this that no one else in the techtube industry does. Steve and Tim at Hardware Unboxed come close a lot of the time, but GN will always be king of kings because of TechJesus!
Dude I just want to say how much I appreciate the depth and quality of GNs testing and methodology. In a climate of surface level, entertainment focused content, you guys shine so bright and I appreciate the hell out of your work.
His point at the beginning reminds me of this effect in people where they’ll read news about something they’re an expert in (like cpu coolers) and see just how badly the media gets it wrong or outright lies, then they’ll turn the page to something they’re not an expert in and be like “oh how interesting, I never knew that” forgetting how bad the last page was.
95% of the time, when you have an armchair expert who thinks they've caught "the media" lying, or even misreporting, it's a better example of the dunning Kruger effect than it is anything else. It's kindof like seeing penises. If you think you see them everywhere, then chances are, you're the one with the problem.
@@TheBrothergreen Your 2nd and 3rd sentences are not strongly related to the 1st, but stated as if they were. I also disagree with your 1st sentence, there are many examples of media/news outlets making overreaching conclusions from scientific papers. In this case it would be the reporters suffering from Dunning Kruger effect. Example: Chocolate is healthy - io9.gizmodo.com/i-fooled-millions-into-thinking-chocolate-helps-weight-1707251800 Your 1st sentence is valid reasoning, however it seems to not be true.
@@alfred0231 "there are many examples of media/news outlets making overreaching conclusions from scientific papers." and there are many, many, many more examples of them A) reporting the science accurately and B) getting someone who knows the science to report the science. Furthermore, this is a cherrypicked situation that does not extrapolate to justify the OP's original assertion. Also, Also, your example fails to even support your premise, let alone the OP's... but no need to dig up another one, I'll readily concede that "the media" is a sensationalist hyperbole machine, but mindlessly repeating the "fake news" mantra is no better than mindlessly believing that a single study proves that chocolate is the cure to weight gain.
TheBrothergreen like just yesterday msnbc said Bloomberg could have given every American $1million dollars instead of spending $500 m on ads.... with 330m people in America that’s like $1.50 per person.
Your dedication to accuracy is unparalleled. As someone who uses an off the shelf mid grade computer to play The Sims and browse memes, I constantly find myself watching your content just because you make it interesting, and go the extra mile to be thorough and precise with your data, and I appreciate that. Good stuff.
Sanya V. Litvyak You haven’t lived until you have built your own system. Even if it’s a budget rig of around $700 USD with a Ryzen 5 2600 and 1660 Super
Actually been waiting for someone outside of the manufacturers testing departments to publicly review without bias, love that many companies get cranky when they are called out too, Carry on GN Jesus.
I think it's funny when anyone gets mad when they are called out for misleading or straight out lies, This process will make the creators more diligent about their products, it can only lead to good things in the end.
THIS is why if I want to know more about a product, first I'll search here on your channel, and if I don't find any, then Google. If you guys have a review, a trust it more than any other media out there. Outstanding work!
Awesome stuff. I remember reading a (probably now defunct) website in like 2005 or so that also used a very similar test methodology with a dummy load for testing coolers, mostly for testing watercooling stuff (I'm talking early stuff like the original Nexxos XP or the 1A-SL2 blocks), but also some phase-change gear. Tower coolers weren't a thing at the time. Even though real-life figures can be interesting as a point of reference, I've always wondered why nobody else had that kind of approach as this is just the best way to approach things and really compare cooling solutions 'apples to apples'. Between this and your power supply test bench, it's refreshing to see a more rigorous 'engineer-y' approach to testing components. As an aside, the only things an end user should really care about rather than actual temperatures are whether their hardware is thermal throttling, and whether the cooler is being subjectively loud.
As a physicist, I really applaud your efforts towards becoming more scientific in your approach to basically everything. Keep up the good work, and keep improving your methods as you discover better ways to do things, that's how to do stuff properly!
Components, Cases and Coolers. This is the best channel for most in depth reviews tbh. You guys really spend a lot of hours in to testing all the stuff. Thanks for all the work you guys do really!
At my old Job i used to do Testing on Contactors, Relays, Switches etc. (worked at a Manufactor of those things). One of the Test's i've done was "Temperature Rise" according to UL and IEC Norms (equip Part with Thermalcouples at all the necessary points and run it on max. Specced Current for a long time and record Time, Amps, Volts, Ambient and every component Temp over a bunch of samples and make fancy graphs and excel sheets with it) I gotta say your testing is very Professional!
So needed and yet wonderful to see finally being established. Looking forward to further details in articles and video 🤩 I apologize for the fact, that I haven't been supporting you financially - until now!
As a former technician from the manufacturing world it is SO nice to see things done right! I know you put a LOT of resources into your testing methodologies in all areas that you test for. IT IS APPRECIATED!! Thank you Steve and crew. I look forward to seeing the results of your labor. JT
This is amazing. I'm shocked at just how much data you have collected. Also how much other videos could be invalid. Thank you for doing so much work. The PC space as a whole thanks you.
Steve and the GN team are literally setting damn industry standards here. In an industry full of BS and “marketing puffery” I’m so glad I can rely on this channel to always get things right. And Steve you have the hardest attitude towards selling out, like even rejecting those executive dinners at tech events. It’s incredible! Love the content!
There's not a single place online where anyone goes into the massive thoroughness that you do. As an enthusiast who loves drilling into the details, it's greatly appreciated.
Probably up there as one of your best videos to date. This was laid out and explained in a very detailed way, and a very good job. I'm looking forward to up coming Air Cooling reviews. Thanks for the detailed facts to back up the content.
Your frustration with this topic is clear to me. One of the very first things you learn as a scientist, is that your data is only as good as your controls - if your controls are flawed, the data is not valid, period. You just spent three years learning this the hard way. *Always ask yourself "What am I really testing here?"* I never believe cooler reviews because they have no controls. Your test bench should provide a method to test the coolers ability to dissipate heat in a controlled manner - this is the *primary* rating of a cooler. BTW, the heater blocks from in-situ PCR thermocyclers can be had cheap and used (it is mostly obsolete tech not), and can provide a simple method of applying pin point heat to imitate the mapping of hot spots in the a CPU. This test though would not test the ability of the cooler to dissipate heat, it would test the ability of the mount to transfer the heat to the pipes.
@mister.T Jr save that tesla would regularly base his claims for something on the rare occasion it worked in the best way he hoped. Others couldn't regularly repeat his outcomes.
David Whitfield Okay but he does another test too... I can’t remember if it’s Max or Auto out-of-the-box but you’re literally complaining about one chart when every video has at least 3 charts and these days it’s often 5+
The power consumption, noise normalization and fan variance have driven me up a wall watching most reviews. I'm simply not equipped to come up with the protocols to properly incorporate them into testing, nor do I have access to the hardware that would be required to do so. Very stoked to be able to benefit from all your hard work and I feel that as with case testing, your cooler testing will be the gold standard in the hardware journalism space. Excited to see the content you have on the way!
The day has finally come! I am very excited to take part of your results going forward. This might not be too sexy content, but bloody hell is it necessary or is it.. Literally noone has tested coolers like it matters, ever. Kudos in advance, and I'll impatiently await the testing. I'll stop by Patreon while I wait.
As a retired aerospace engineer who was involved with thermal testing (among many other types of testing), I would not even consider using the CPU as part of the test strategy, there are simply too many variables. I'm only interested in the data from your H/W test fixture where you have much tighter control of the thermal load placed on the UUT. IMO, this gives me a much better idea of which products perform the best relative to the others.
You know, if you ever head back to school (University), the processes you're coming up with and the data you've collected would make a really good PhD thesis.
you usually collect large amounts of data and ensure the quality of that data is accurate when conducting a study. the data here is only interesting for consumer reasons and brand integrity, wouldnt say it would be worth much in school but the way he conducts the data collection shows that he did ok in school at least.
Like several others here, this video made me decide to become a patreon - SUPERB work. I come from an R&D Engineering background (IC engines of all types/sizes) and know how important repeatability is. To be able to control things is key, so that box you had made, made me sit up, as it is the only way to control temps, repeatably and accurately. Fantasic work Steve.
This! This is why I respect nearly every single thought and opinion (and fact) that comes from this channel. You guys go above and beyond to not just deliver accurate data and results but go above and beyond to make sure we, the audience, understand the methods used. Please know that it is tremendously appreciated...
8:55 if you need to use the same exact hardware for every test, what happens when a component dies or is damaged? How do you compensate for the differences of a new part? Do you just retest coolers and then just factor in the difference?
Step 1: Hope it doesn't die Step 2: Buy a box of replacements and individually test them all against known data, then see if any of them replicate the results within an acceptable margin Step 3: Adjust margins if so (for example, maybe you go from +/- 0.5C to +/- 1.0C), otherwise, begin arduous process of retesting everything!
Gamers Nexus Thank you for taking the time to answer my questions. (Sorry there’s so many). If you guys aren’t planning to, maybe I can go through some comments before there’s too many and put together an FAQ of viewer questions you answer.
@@cocaine_is_a_hell_of_drug765 I'm curious about that too but I imagine they'll end up doing sanity check tests with known hardware/coolers and compare with old data to account for discrepancies and degradation
@@cocaine_is_a_hell_of_drug765 how hardware degradation (what in particular you're referring?) would end up in different temperatures? Without changing consumed power, because it's measured/being monitored?
Just amazing! You say that this content isn't exciting but it really is. I don't see anyone else putting in the amount of effort that everyone there has to properly benchmark coolers. The key word there is "properly"! I can't wait to see the review videos and know that I can finally trust the results.
Steve, you and your team really deserve all my respect. The methdology you put into testing, and this particular topic which has so many variables, and in the end usually translates into perceived performance is essential not only for us nerds and geeks who are passionate about PCs, but also to manufacturers of both the products you test and feature, and for those doing system's integration and building. The degree (pun intended) of attention to details so to avoid the dreaded bias, with emphasis in reproducibility and automation to the greatest extent and the lengths you have gone in order to produce such... Hat's off to you, Sir. I'm eager to see more content on this topic espcially using that "dummy heater", which is... mind blowing (every cooler manufacturer I suppose has their own version of one or similar?). At any rate, kudos. Keep it up, you find ways to amaze us every time you post out a new video (even if it is only to tell the "news") PS: I bet many universities will "borrow" methodology from your tests (and do not mean thermals only, but in general).
I’m glad someone else feels my pain. I did a lot of thermal resistance testing and ended up testing cooler designs at ~10 different power levels, dealing with linear regressions and other fun aspects of cooler testing.
Just use the hot plate to compare them. I don’t think people care about real world because even if you make the same mistakes with different coolers in the real world, you should still get a similar difference compared to a perfect scientific test using the hot plate
One rarely (mostly never) reported set of numbers is a coolers saturation curve. Most tests are done with at most three data points, idle, stock and overclocked. Each point representing a wattage range through inconsistencies. I've seen results where a 96mm three pipe cooler outperformed a popular 4 pipe 120mm cooler, but the references stopped at full load stock (95W). What wasn't considered was the saturation points of the coolers and the fact that after that point the 96mm three pipe cooler turned to shite thereafter. Each cooler will have a thermal curve that will vary depending on the wattage being dissipated, a useful graph would be data points of performance in the temperature delta at an even rise in wattage, e.g idle 65W +20W per data point up to saturation or on up to 265W if capable. Coolers that perform closely should show variance at each data point, this gives the end user information that gets close to their particular use case as they can see the performance at the wattage they're likely to peak at, be that 95W, 125W .... or 180W or greater. It'd also make a nice comparative line graph from idle to insane power draws.
Great stuff, nice to see such a comprehensive methodology and a test box again, haven't really seen one since the early 2000's before the days of youtube from a site long gone...Back then we were using car heater radiators with custom made shrouds and blocks here in Greece at least. I used such a setup to cool a thoroughbred B back in the day, it was so funny the petshop employee's face every time i went and asked to buy a new water pump (they broke often back then) the employee would ask the aquariums size and i was like no it's to cool a computer. I would like to see a mugen 5 tested its a great vfm cooler with very good performance.
guessing based on bad memory here but... pretty sure those work worse than paste so itd end up looking worse and not being somewhat similar to what a consumer would be getting
I knew you guys put a lot of effort in testing, but I didn't think it was this deep and precise. This team is just the best of the best! Also, it's amazing that you share all this methodology and techniques with everyone, thanks a lot!
And you're still going off BS numbers. Unless you have Steve come to your house and measure the temperature you're getting. Because in the end that's all that really matters.
It's impossible for a CPU cooler review to be completely accurate. To get completely accurate results, you have to test on every compatible CPU, on every compatible motherboard, with every compatible UEFI/BIOS release, in every enclosure, with every possible fan configuration. Throwing a CPU cooler on an open air test bench, produces inaccurate results. Throwing a CPU cooler on a stand-alone heater, produces inaccurate results. Removing variables, does not produce accurate results, because those variables have a direct affect on the final data set.....which is the data set created by the user in real-world usage. Just because a CPU cooler produces the best results on a stand-alone heater, doesn't mean it will be capable of adequately cooling a CPU in any particular enclosure. If you want real-world applicable results, you test in real-world scenarios. Trying to control every single possible aspect of a test, does not produce accurate results. In real world usage, voltages fluctuate. In real-world usage, fan RPM fluctuates. In real-world usage, temperatures fluctuate. By trying to remove such variables, your data sets become useless. Most of us aren't looking for "perfect world" test results. We want "real-world" applicable results. To that end, I'll continue looking at reviews from sites and channels that do "real-world" oriented testing.
@@aaronthomas6155 introducing variables is not the way to increase test accuracy. CPUs themselves are not very reliable heating elements either. We know they make heat. There's just too many variables that determine how much heat they dissipate though. You're way better off with a simple heating element if you want to measure cooler heat dissipation.
This is arguable the closest I've seen anyone get to a college level test methodology breakdown outside of a actual school. I've listen to many college lectures and this is one of the few dissertations that made my brain hurt just trying to process it all. If some of your videos aren't used in computer classes i would be surprised. Well advanced classes, with tech quickie for the beginner cause yours take some prior knowledge to process sometimes.
This is why I watch your channel. You don’t stop at the status quo, even though it’s easier but you actually try and understand and figure out the little nuances so that you can give us the most complete picture possible. Maybe CM can hook GN up with one of those fancy cooler test chambers
I do get what you are saying in the video and I applaud you guys for doing your very best to make sure every aspect of cooler testing is done and done right. I do feel though that not only do coolers need to be tested in a 100% controlled test. I also feel every cooler also needs to be tested as before on actual boards with CPU's to see how the cooler.CPU,main board all work together in a real computer setup to get that data as well. I also see the need for the Delta over amb tests but I also feel actual temps of the CPU should be shown in the graphs mainly because we then get both sets of data to look at. Also if Amb temp is not shown in a cooler review and the tests only show Delta over Amb with have no idea what the CPU was actually running at under full load tests. In my own system with to tests temps I use the same software every time to test my CPU temps under full CPU load. The biggest thing that can change my CPU temps is amb temp. So if my amb temp is 21c my CPU may go up to 68c as a max temp after 30 minutes of full load using a CPU torture test software and yet in the same 21c amb temp in a game the CPU only runs at maybe 47c-49c in a game load. Then if amb temp goes up to 28c in the summer months that same test will raise to 75c under full load torture test. In game temps also raise to 54c-56c in the same amb temp of 28c. So in my example temps go up as expected with amb temp increase to pretty close or right on the money. I use a 280mm EVGA CLC on my CPU and noticed the temps pretty much flatline after about 20 minutes as the CLC reaches it working temp anything after that seems to pretty hold pretty steady in very long game sessions or full load torture tests. My CPU is a old i7 2600K@5.1GHz with controlled voltage set to max out at 1.47 volts with load line set to turbo on my Gigabyte board if I set it to extreme load line the board then tries to over shoot my set max voltage. I used to use a big air cooler on this CPU form 2011-2018 and was only able to get 5.0GHz before the cooler would get to hot and 5.1GHz was a no go and required more voltage. After switching to the 280mm CLC I was able to lower voltage down to 1.45 volts for 5.0GHz and 1.47 @5.1GHz was finally a go and have good temps. I feel my testing on my own system has shown me that everything is pretty much spot on and is 100% rock solid stable for such an old platform in 2020.
For a moment there, I thought you had a box that could theoretically show cooling performance, from any given heatsink and or cooler based on the TDP specifications of computer processors!
I worked for a large American semi-conductor company as a software engineer supporting test teams, particularly on the statistical process control and process capability side, and seeing your methodology is pretty impressive. Maybe the best I've seen from an "amateur" setup (meaning one without millions of dollars in testing equipment). People probably don't realize the number of factors for which you have to control to get valid results and the amount of variance (and RMA's) you'll get if you don't properly control those variables. Impressed!
"People want to see real world data" Sure, I want to see it validated against real world data, but I'm much more interested in that heater and the numbers it would produce. I'd trust that far more than seeing "real world data"
Lets see, we have Hardware Unboxed for Monitors, Actually Hardcore Overclocking for Motherboards, JayzTwoCents for Water Cooling, and Gamers Nexus for Cases and CPU coolers. CPU and GPU data is done fairly universally well from a lot of the big names (Your a big name to me GN =D). Once you get your crazy setup for Power Supplies and case fans done, that is basically the whole computer. I am more then happy to continue supporting your guys through the GN Store and Patreon with the amount of passion and integrity that you bring to your work. Thank you GN Team.
Not gana lie, I really liked seeing and learning about the extensive measures you guys go through to ensure proper and accurate data on such a volatile and difficult thing. I've only fussed a little with doing my own testing and realized there was so many variables, so thank you and good job taking the time to break it down and work out how to control them as best as humanly possible! Keep up the good fight against misinformation!
After watching your vids for almost a year now, this was the one that made me subscribe. Excellent content and I'm impressed with the effort you've put in to reliably and accurately measure cooler performance. Worthy of a doctorate - well done!
I used to really dislike sites that used a static thermal testload for cooler reviews. You have done an excellent job of reversing my opinion... hope to be seeing static load cooler results in the future!
Smashing the like button so hard. The past week I was looking for an air cooler and went through your videos, web page and more and after that I finally gave up as your numbers are the only ones I would trust for something like this 😅 Looking so much forward to seeing the high end air coolers like NH-D15,, 15S, D14 Dark rock pro 4 etc. battle it out. For some reason I find your thermal testing interesting and almost soothing, even if I don't need e.g. a case or similar, keep up the good work 😊
While it can be reused, can it be reused reliably? What happens after it's been clamped and removed 100 times? Is it still just as conductive? What about clamping pressure? Are they consistent with different coolers that clamp differently? It'd introduce more variables, and most users will be using paste, so the best bet is to just use a known good paste with a known good application method
Paste has a consistently better thermal transfer so differences would be more pronounced and recordable. This is a drastic analogy but it would be like putting a/c in a closed off room and measuring how the temperature changes is the room next to it. There’s probably going to be a difference but it would be a lot easier to measure if there were vents.
Man, THANK YOU for this. This was a joy from start to finish seeing how motivated, focused, and most importantly passionate about tech and accuracy you guys are. I can't wait to see more cooler stuff
Gamers Nexus in 2069: And that is why as a reviewer you need to control what you eat, this can invalidate testing because eating foods with high calorific makes your breathing rate increase and the hot air given out is enough to completely invalidate a cooler's performance. We, at Gamers Nexus, have completely standardised what our technicians eat due to this exact issue. Also, all our technicians are required to maintain a constant 15 breaths per minute from test to test so as to reduce variance in that respect.
Hats off to all of you at GamersNexus for setting the gold standard for testing and reviews. Nobody else I can think of comes anywhere close to the level of accuracy and legitimacy you guys provide and strive for. Keep up the great work!
Invalid results. You failed to take into account: 1. The current Moon phase 2. Your cat's voltage level 3. Whether you have your hair in front of, or behind your shoulder(s) 4. My feelings 5. The median price of chicken tenders I expected better Steve.
There's nothing else I can say but BRAVO! You could not have done this and the lion's share of us wouldn't know the difference, but this is so above and beyond that it truly elevates and sets a new standard. What you all provide is an incredible service and you deserve all the accolades you get. Simply stellar work!
I hope we get a good PSU testing as well in this decade.... based on how long/much this refined... Also, since now thermal-paste testing is awaible as well--- i guess...
Thanks for paying attention to the details and knowing what those details are. Things like knowing that CPUs within the same SKU and run are going to have a +/- 10% difference from chip to chip is why I keep coming back to GN and don't mind spending a bit extra when I need stuff you carry.
Please support our reinvestment into GamersNexus and its testing quality and accuracy via Patreon (www.patreon.com/gamersnexus ) or the GN Store (store.gamersnexus.net )
Article accompaniment (the video and article both have some unique portions): www.gamersnexus.net/guides/3561-cpu-cooler-testing-methodology-most-tests-are-flawed
Hey Steve, does your merch store deliver those mod mats to Mexico? I need a couple of them.
@@BrawndoQC We ship internationally!
When you include meow meow cooling information for all your cooler reviews I'll sub to patreon
Any ETA on the GN Toolkit restock?
With the emphasis on using the exact same hardware on every test and the demounting and remounting of the hardware so many times what are your concerns about breaking pins on the CPU?
and this is why I decided not to do thermal testing
I think I understand what you meant now...
Coming from Buildzoid, this is a big compliment. "I cannot fucking believe you went and did it, but you did it."
“Guy who performs extreme RAM overclocks states thermal testing isn’t worth it”, thus, the reason for which the cooler testing scene is comparatively poor
But you dooooo. Margin of error 1C meh good enough. XD
Well it either works or if doesn't, if it melts you probably went too far. 🤣
6:30 "And expect a lot of cooler content in the next couple of weeks"
I mean, your content is pretty cool already imo
🙄
*audible groan*
Cute...
Are you sure you used the correct methodology to measure the coolness of their content?
@@djpob GN approved coolness methods are pretty hard to find
I just want to say that I personally appreciate immensely the lengths to which you guys go to produce actual, reproducible and accurate data. In the last month I've looked at 6 reviews of a cooler, 1 of which put it ahead of the D15, 1 of which put it behind the D15, and 4 of which put it at about equal to the D15. This is VERY CLEARLY NOT A SIMPLE THING TO TEST FOR, and I applaud you for taking the Noctua approach.
By which I mean: Noctua hasn't put out a new top of the line cooler since 2014, because they didn't feel like they could scientifically justify that they'd come up with an ACTUAL upgrade. The marketing department could have EASILY come up with at least 4 new coolers in the meantime, much like how many other manufacturers have. Now whether you actually buy Noctua's claims or not, the fact that they haven't introduced a new top of the line cooler since 2014 tells you something about how much they believe in their OWN product. (How many other companies outright tell you that they've worked on a fan design for 4 years, and are going to need at least 2 years more to make it perfect?)
And you originally said you'd come back to cooler reviews years ago, yet didn't actually feel comfortable enough to do it until now, years later, tells me that you're not just trying to pull in views. This to me says that you want to actually do things right, rather than just push out numbers to boost your views. And for that I thank you.
1000000% this.
THIS is why I trust GamersNexus above all else when it comes to hardware... Making actually reliable, reproducible data for something is NOT easy... and data quality is just not a thing you can hand wave to create. It takes a ton of know-how to do it right.
Unfortunately this testing still does not account for die size and location.
@@Mr.Morden reading in between the lines, I assume it will be! Other it would be a shame - but I doubt it.
@@Mr.Morden well that is accounted with the test where GN does the cooler mounting pressure and contact .
And if the cooler is mounted on Intel or and I understand the die size and location are different interms of monolithic and chiplet design but the coolers are not made Intel or AMD specific they are compatible
AMEN
A large part of my job is testing equipment for the military. I feel like I just had my job explained me by someone who doesn't work there, and still understands it way better than me. Outstanding explanation.
Glad to hear that the same concerns of not trusting any software or hardware persists in all industries!
Oddly if you want to find better methods then talk to someone who is in the fields of medical or food science based lab work studying organisms like listeria monocytogenes. They have to control temperatures quite accurately.
I work in research, this is why it is so necessary to know the test setup when evaluating results. Anytime anyone presents me with "new research indicates x" it always brings skepticism on my part. Good job Gamers Nexus!
These guys should be contracted. I mean my god. I have been a viewer for a long time. Sure they joke but no one else on the planet goes into this lvl of depth on a consistent basis.
@@GamersNexus it seems as though y'all are starting the deep dive into metrology. Welcome to the depths.
Just don't see this level of professionalism anywhere else.
try studying physics, this brings me back to those days
rtings.com does amazing testing as well. Their TV testing is miles beyond anyone else. Unfortunately not a lot of PC stuff yet (they are just starting on keyboards and mice). It would be awesome if Gamers Nexus could create a similar exhaustive database of cooler and case reviews (and more, certainly), I'm sure everyone in the industry will send them their stuff.
and techpowerup is decent.
@@deminybs agreed
Physics, real engineering, medical lab work, food science research into micro organisms. There's lots of professionalism out there, it just seems to end with PC manufacturers.
Gamers Nexus year 2050: "We now have complete control of weather globably so external temps should be more steady and while not perfect the data should be a bit more accurate"
I have a feeling that GN will be the company that most other companies and channels will start to go to for the true stats of products moving forward. It could even be a situation when people do reviews that they use a term like "Gamers Nexus Testings has found ...... " to report true validity in the numbers they cite.
Super in depth and obviously alot of work goes into it.
You say that like it isn't already. Between GN and Buildzoid, this is the place I go and send others when I/they have a question.
@@JamesErath You don't see companies like Nvidia, Intel, AMD, TheVerge, LMG, etc saying "Go to buildzoid and/or GN", or "GN found this:", which is what hes saying. Yes WE say go to GN or buildzoid, but companies themselves do not
Theres so much fake news out there. Thats why I dont trust Jay in temps because of where hes from.
Except Tom's Hardware, who will do an obviously invalid test and then call GN liars and trolls.
@@mjc0961 That's Hardware Unboxed, not GN
1:37 As a 25-year QA professional, this statement is accurate just about everywhere. Especially in software. Trust, but verify. Log everything. And remember: the translation programs work *FOR* the Matrix. ;)
I've done software QA for 1 year, and yeah. Everything behaves differently all the time, on every machine, and the same machine.
God it could annoy the devs :D
Pretty damn superb. You might just be the first creator to win patreon dollars from me. Not for my own benefit. I kind of own the stuff I need to own and won't be buying for a few years. But this content is invaluable to new members of the community, and that I support wholly.
Just never buy a teleprompter K. Ad hoc/page shuffling Steve is best Steve.
same here i never considered patreon until this video. i will pay for this kind of content
Ha! I feel like the page shuffling helps me think a lot. We tried a teleprompter before and it really didn't work well with how I go off script!
Steve and crew were the first Patreon support I ever offered to creators. Been a proud Patreon for more than 2 years now, and it's exactly because of content like this that no one else in the techtube industry does. Steve and Tim at Hardware Unboxed come close a lot of the time, but GN will always be king of kings because of TechJesus!
Tim is great for monitor reviews
Dude I just want to say how much I appreciate the depth and quality of GNs testing and methodology. In a climate of surface level, entertainment focused content, you guys shine so bright and I appreciate the hell out of your work.
I just love the dedication of Tech Jesus and his crew. I don’t even think some manufacturers do as much or accurate testing.
His point at the beginning reminds me of this effect in people where they’ll read news about something they’re an expert in (like cpu coolers) and see just how badly the media gets it wrong or outright lies, then they’ll turn the page to something they’re not an expert in and be like “oh how interesting, I never knew that” forgetting how bad the last page was.
95% of the time, when you have an armchair expert who thinks they've caught "the media" lying, or even misreporting, it's a better example of the dunning Kruger effect than it is anything else. It's kindof like seeing penises. If you think you see them everywhere, then chances are, you're the one with the problem.
@@TheBrothergreen Your 2nd and 3rd sentences are not strongly related to the 1st, but stated as if they were.
I also disagree with your 1st sentence, there are many examples of media/news outlets making overreaching conclusions from scientific papers. In this case it would be the reporters suffering from Dunning Kruger effect.
Example: Chocolate is healthy - io9.gizmodo.com/i-fooled-millions-into-thinking-chocolate-helps-weight-1707251800
Your 1st sentence is valid reasoning, however it seems to not be true.
@@alfred0231 "there are many examples of media/news outlets making overreaching conclusions from scientific papers."
and there are many, many, many more examples of them A) reporting the science accurately and B) getting someone who knows the science to report the science.
Furthermore, this is a cherrypicked situation that does not extrapolate to justify the OP's original assertion.
Also, Also, your example fails to even support your premise, let alone the OP's... but no need to dig up another one, I'll readily concede that "the media" is a sensationalist hyperbole machine, but mindlessly repeating the "fake news" mantra is no better than mindlessly believing that a single study proves that chocolate is the cure to weight gain.
TheBrothergreen
Maybe in some specific areas, but with general news and politics, absolutely not. Journalist make massive errors all the time.
TheBrothergreen like just yesterday msnbc said Bloomberg could have given every American $1million dollars instead of spending $500 m on ads.... with 330m people in America that’s like $1.50 per person.
Your dedication to accuracy is unparalleled. As someone who uses an off the shelf mid grade computer to play The Sims and browse memes, I constantly find myself watching your content just because you make it interesting, and go the extra mile to be thorough and precise with your data, and I appreciate that. Good stuff.
Sanya V. Litvyak You haven’t lived until you have built your own system. Even if it’s a budget rig of around $700 USD with a Ryzen 5 2600 and 1660 Super
Actually been waiting for someone outside of the manufacturers testing departments to publicly review without bias, love that many companies get cranky when they are called out too,
Carry on GN Jesus.
None of this means beans. The only thing that matters is the temperature you can manage to achieve.
I'm surprised Consumer Reports didn't tackle this years ago.
I think it's funny when anyone gets mad when they are called out for misleading or straight out lies,
This process will make the creators more diligent about their products, it can only lead to good things in the end.
THIS is why if I want to know more about a product, first I'll search here on your channel, and if I don't find any, then Google. If you guys have a review, a trust it more than any other media out there.
Outstanding work!
Steve’s frustration on this topic really cuts through in this video 😂
Awesome stuff. I remember reading a (probably now defunct) website in like 2005 or so that also used a very similar test methodology with a dummy load for testing coolers, mostly for testing watercooling stuff (I'm talking early stuff like the original Nexxos XP or the 1A-SL2 blocks), but also some phase-change gear. Tower coolers weren't a thing at the time.
Even though real-life figures can be interesting as a point of reference, I've always wondered why nobody else had that kind of approach as this is just the best way to approach things and really compare cooling solutions 'apples to apples'.
Between this and your power supply test bench, it's refreshing to see a more rigorous 'engineer-y' approach to testing components.
As an aside, the only things an end user should really care about rather than actual temperatures are whether their hardware is thermal throttling, and whether the cooler is being subjectively loud.
Random yt channel: "so, here's the results of our testing of this CPU cooler"
GN Steve: "You are a fool and a waste of time"
As a physicist, I really applaud your efforts towards becoming more scientific in your approach to basically everything. Keep up the good work, and keep improving your methods as you discover better ways to do things, that's how to do stuff properly!
I can't wait to see more about that heater. --- Words I'd never think I'd say.
Components, Cases and Coolers. This is the best channel for most in depth reviews tbh. You guys really spend a lot of hours in to testing all the stuff.
Thanks for all the work you guys do really!
18:06 STRICKEN
+1 for verbiage that peaked in usage around 1900
I really love these methodology explanation vids. You guys are so thorough.
At my old Job i used to do Testing on Contactors, Relays, Switches etc. (worked at a Manufactor of those things). One of the Test's i've done was "Temperature Rise" according to UL and IEC Norms (equip Part with Thermalcouples at all the necessary points and run it on max. Specced Current for a long time and record Time, Amps, Volts, Ambient and every component Temp over a bunch of samples and make fancy graphs and excel sheets with it)
I gotta say your testing is very Professional!
cant wait to see you test the IceGiant as was shown on ltt when the production model comes to market
We have requested a production unit! We decided not to work with the prototype and to just wait for the final product for review.
@@GamersNexus I'm hoping it lives up to the claims because that would be awesome but I think it might be a gimmick
So needed and yet wonderful to see finally being established. Looking forward to further details in articles and video 🤩
I apologize for the fact, that I haven't been supporting you financially - until now!
You are about to make Tom's Hardware very upset, they might even call you a liar and a troll for something like this.
Really, I figured they where reasonably trustworthy?
@@DeBerjeuh they have recently been thrown into the "iffy, keep an eye on them" category.
wait until you see Tom's Software
How could anyone ever forget Tom's "Just Buy It!" Hardware.
@@NordicGapper exactly!
As a former technician from the manufacturing world it is SO nice to see things done right! I know you put a LOT of resources into your testing methodologies in all areas that you test for. IT IS APPRECIATED!! Thank you Steve and crew. I look forward to seeing the results of your labor.
JT
I trust Noctua more than myself.
I trust them to make the ugliest fans in existence. Good thing I never look inside my PC. :P
I fail all the time, but Noctua has never failed me
Noctua, the only company to make beige and brown a status symbol in a looks and RGB oriented industry.
John They make black ones now
@@RAYSGT why do you need to paint them, they released chroma black versions now the prices aren't that much diff
This is amazing. I'm shocked at just how much data you have collected. Also how much other videos could be invalid. Thank you for doing so much work. The PC space as a whole thanks you.
Steve and the GN team are literally setting damn industry standards here. In an industry full of BS and “marketing puffery” I’m so glad I can rely on this channel to always get things right. And Steve you have the hardest attitude towards selling out, like even rejecting those executive dinners at tech events. It’s incredible! Love the content!
There's not a single place online where anyone goes into the massive thoroughness that you do. As an enthusiast who loves drilling into the details, it's greatly appreciated.
I watched this entire thing and really I'm just blown away at basically inventing cooler standard testing
Probably up there as one of your best videos to date. This was laid out and explained in a very detailed way, and a very good job. I'm looking forward to up coming Air Cooling reviews. Thanks for the detailed facts to back up the content.
I'm so glad there is someone on the tech community that goes to all this trouble to produce proper data, kudos to you guys!
Your frustration with this topic is clear to me.
One of the very first things you learn as a scientist, is that your data is only as good as your controls - if your controls are flawed, the data is not valid, period.
You just spent three years learning this the hard way. *Always ask yourself "What am I really testing here?"*
I never believe cooler reviews because they have no controls.
Your test bench should provide a method to test the coolers ability to dissipate heat in a controlled manner - this is the *primary* rating of a cooler.
BTW, the heater blocks from in-situ PCR thermocyclers can be had cheap and used (it is mostly obsolete tech not), and can provide a simple method of applying pin point heat to imitate the mapping of hot spots in the a CPU. This test though would not test the ability of the cooler to dissipate heat, it would test the ability of the mount to transfer the heat to the pipes.
This should be called "Why our data is good and everyone else's sucks."
"Why all data sucks, but ours sucks less than yours"
This should be called "Why our data is made up in a lab and no one can verify our conclusions".
@@levyroth that's actually just regular academic research
@mister.T Jr save that tesla would regularly base his claims for something on the rare occasion it worked in the best way he hoped. Others couldn't regularly repeat his outcomes.
David Whitfield Okay but he does another test too... I can’t remember if it’s Max or Auto out-of-the-box but you’re literally complaining about one chart when every video has at least 3 charts and these days it’s often 5+
The power consumption, noise normalization and fan variance have driven me up a wall watching most reviews. I'm simply not equipped to come up with the protocols to properly incorporate them into testing, nor do I have access to the hardware that would be required to do so. Very stoked to be able to benefit from all your hard work and I feel that as with case testing, your cooler testing will be the gold standard in the hardware journalism space. Excited to see the content you have on the way!
The day has finally come! I am very excited to take part of your results going forward.
This might not be too sexy content, but bloody hell is it necessary or is it.. Literally noone has tested coolers like it matters, ever.
Kudos in advance, and I'll impatiently await the testing. I'll stop by Patreon while I wait.
As a retired aerospace engineer who was involved with thermal testing (among many other types of testing), I would not even consider using the CPU as part of the test strategy, there are simply too many variables. I'm only interested in the data from your H/W test fixture where you have much tighter control of the thermal load placed on the UUT. IMO, this gives me a much better idea of which products perform the best relative to the others.
This sort of stuff is my favourite content. So while I may be a minority of your viewers, I really appreciate videos like this.
You accuracy and meticulous attention to detail is astounding. Great work!
You know, if you ever head back to school (University), the processes you're coming up with and the data you've collected would make a really good PhD thesis.
nommindymple school! Your joking right?
you usually collect large amounts of data and ensure the quality of that data is accurate when conducting a study. the data here is only interesting for consumer reasons and brand integrity, wouldnt say it would be worth much in school but the way he conducts the data collection shows that he did ok in school at least.
10 / 10 comment, I don't really care for merch but I would subscribe to Patreon and buy shirts and mod mats if it meant Steve getting a PhD
Like several others here, this video made me decide to become a patreon - SUPERB work. I come from an R&D Engineering background (IC engines of all types/sizes) and know how important repeatability is. To be able to control things is key, so that box you had made, made me sit up, as it is the only way to control temps, repeatably and accurately. Fantasic work Steve.
This is exactly why i love Gamers Nexus. So thorough. Much science!
This! This is why I respect nearly every single thought and opinion (and fact) that comes from this channel. You guys go above and beyond to not just deliver accurate data and results but go above and beyond to make sure we, the audience, understand the methods used. Please know that it is tremendously appreciated...
8:55 if you need to use the same exact hardware for every test, what happens when a component dies or is damaged?
How do you compensate for the differences of a new part? Do you just retest coolers and then just factor in the difference?
Step 1: Hope it doesn't die
Step 2: Buy a box of replacements and individually test them all against known data, then see if any of them replicate the results within an acceptable margin
Step 3: Adjust margins if so (for example, maybe you go from +/- 0.5C to +/- 1.0C), otherwise, begin arduous process of retesting everything!
Gamers Nexus
Thank you for taking the time to answer my questions. (Sorry there’s so many).
If you guys aren’t planning to, maybe I can go through some comments before there’s too many and put together an FAQ of viewer questions you answer.
@@GamersNexus
How do you account for hardware degradation over time?
@@cocaine_is_a_hell_of_drug765 I'm curious about that too but I imagine they'll end up doing sanity check tests with known hardware/coolers and compare with old data to account for discrepancies and degradation
@@cocaine_is_a_hell_of_drug765 how hardware degradation (what in particular you're referring?) would end up in different temperatures? Without changing consumed power, because it's measured/being monitored?
love to see how ethical you guys are on your reviews and how nothing is left to chance
keep up the good work
I trust Gamers Nexus...
What? You mean not 'JUST BUY IT' ? :)
@El DIABLO every GPU review I've seen on this channel has been very in depth
Trust no one.
@@Aspartame69 low quality bait. Try again
@@Aspartame69 Your dislike of GN (and their audience) is very evident, so why are you here watching and commenting on their videos?
Just amazing! You say that this content isn't exciting but it really is. I don't see anyone else putting in the amount of effort that everyone there has to properly benchmark coolers. The key word there is "properly"! I can't wait to see the review videos and know that I can finally trust the results.
"We want to consistently test CPU coolers"
Windows 10 update has entered the chat:
Steve, you and your team really deserve all my respect.
The methdology you put into testing, and this particular topic which has so many variables, and in the end usually translates into perceived performance is essential not only for us nerds and geeks who are passionate about PCs, but also to manufacturers of both the products you test and feature, and for those doing system's integration and building.
The degree (pun intended) of attention to details so to avoid the dreaded bias, with emphasis in reproducibility and automation to the greatest extent and the lengths you have gone in order to produce such... Hat's off to you, Sir. I'm eager to see more content on this topic espcially using that "dummy heater", which is... mind blowing (every cooler manufacturer I suppose has their own version of one or similar?). At any rate, kudos.
Keep it up, you find ways to amaze us every time you post out a new video (even if it is only to tell the "news")
PS: I bet many universities will "borrow" methodology from your tests (and do not mean thermals only, but in general).
I’m glad someone else feels my pain. I did a lot of thermal resistance testing and ended up testing cooler designs at ~10 different power levels, dealing with linear regressions and other fun aspects of cooler testing.
definitely RESPECT for the depth you all go for the sake of proper testing.
thank you. For real!
Just use the hot plate to compare them. I don’t think people care about real world because even if you make the same mistakes with different coolers in the real world, you should still get a similar difference compared to a perfect scientific test using the hot plate
One rarely (mostly never) reported set of numbers is a coolers saturation curve. Most tests are done with at most three data points, idle, stock and overclocked. Each point representing a wattage range through inconsistencies. I've seen results where a 96mm three pipe cooler outperformed a popular 4 pipe 120mm cooler, but the references stopped at full load stock (95W). What wasn't considered was the saturation points of the coolers and the fact that after that point the 96mm three pipe cooler turned to shite thereafter.
Each cooler will have a thermal curve that will vary depending on the wattage being dissipated, a useful graph would be data points of performance in the temperature delta at an even rise in wattage, e.g idle 65W +20W per data point up to saturation or on up to 265W if capable. Coolers that perform closely should show variance at each data point, this gives the end user information that gets close to their particular use case as they can see the performance at the wattage they're likely to peak at, be that 95W, 125W .... or 180W or greater. It'd also make a nice comparative line graph from idle to insane power draws.
"I just wanted to know which cooler you recommend, man..."
Great stuff, nice to see such a comprehensive methodology and a test box again, haven't really seen one since the early 2000's before the days of youtube from a site long gone...Back then we were using car heater radiators with custom made shrouds and blocks here in Greece at least. I used such a setup to cool a thoroughbred B back in the day, it was so funny the petshop employee's face every time i went and asked to buy a new water pump (they broke often back then) the employee would ask the aquariums size and i was like no it's to cool a computer.
I would like to see a mugen 5 tested its a great vfm cooler with very good performance.
Ever tried graphite thermal pads instead of thermal paste? For more constant repeated mounting of coolers.
guessing based on bad memory here but...
pretty sure those work worse than paste so itd end up looking worse and not being somewhat similar to what a consumer would be getting
I knew you guys put a lot of effort in testing, but I didn't think it was this deep and precise. This team is just the best of the best!
Also, it's amazing that you share all this methodology and techniques with everyone, thanks a lot!
I was suprised u never tested a scythe cooler?! Would love to see a test from you; especially as it is not trivial!
Keep up the good work guys!
Love the content! Don't give up!
I wish we had this back in the early '00s when things started to get crazy. We've been going off of BS numbers for 20 years.
And you're still going off BS numbers. Unless you have Steve come to your house and measure the temperature you're getting. Because in the end that's all that really matters.
It's impossible for a CPU cooler review to be completely accurate. To get completely accurate results, you have to test on every compatible CPU, on every compatible motherboard, with every compatible UEFI/BIOS release, in every enclosure, with every possible fan configuration. Throwing a CPU cooler on an open air test bench, produces inaccurate results. Throwing a CPU cooler on a stand-alone heater, produces inaccurate results. Removing variables, does not produce accurate results, because those variables have a direct affect on the final data set.....which is the data set created by the user in real-world usage. Just because a CPU cooler produces the best results on a stand-alone heater, doesn't mean it will be capable of adequately cooling a CPU in any particular enclosure.
If you want real-world applicable results, you test in real-world scenarios. Trying to control every single possible aspect of a test, does not produce accurate results. In real world usage, voltages fluctuate. In real-world usage, fan RPM fluctuates. In real-world usage, temperatures fluctuate. By trying to remove such variables, your data sets become useless. Most of us aren't looking for "perfect world" test results. We want "real-world" applicable results. To that end, I'll continue looking at reviews from sites and channels that do "real-world" oriented testing.
@@aaronthomas6155 introducing variables is not the way to increase test accuracy. CPUs themselves are not very reliable heating elements either. We know they make heat. There's just too many variables that determine how much heat they dissipate though. You're way better off with a simple heating element if you want to measure cooler heat dissipation.
You deserve an honorary plaque from NIST for your commitment to testing consistency.
Oh Yea Ive been waiting this for so long Thank you Steve and co. stay as cool as your thermal solutions in this time of epidemic
Thanks! Trip is going great so far. We've been waiting to work on this for years!
This video is.. like a clear answer to "what is this channel about?" :) It's hard to pinpoint a more GN-style video. Love it!
This is arguable the closest I've seen anyone get to a college level test methodology breakdown outside of a actual school. I've listen to many college lectures and this is one of the few dissertations that made my brain hurt just trying to process it all. If some of your videos aren't used in computer classes i would be surprised. Well advanced classes, with tech quickie for the beginner cause yours take some prior knowledge to process sometimes.
glaurung89 same feeling. It totally sounds like he’s reading a physics lab report. Great use of language too.
This is why I watch your channel. You don’t stop at the status quo, even though it’s easier but you actually try and understand and figure out the little nuances so that you can give us the most complete picture possible.
Maybe CM can hook GN up with one of those fancy cooler test chambers
This is why I love GN. I'm a physics student and i feel like even in this field sometimes the testing isn't done to the level you guys do it. Hats off
I do get what you are saying in the video and I applaud you guys for doing your very best to make sure every aspect of cooler testing is done and done right. I do feel though that not only do coolers need to be tested in a 100% controlled test. I also feel every cooler also needs to be tested as before on actual boards with CPU's to see how the cooler.CPU,main board all work together in a real computer setup to get that data as well.
I also see the need for the Delta over amb tests but I also feel actual temps of the CPU should be shown in the graphs mainly because we then get both sets of data to look at. Also if Amb temp is not shown in a cooler review and the tests only show Delta over Amb with have no idea what the CPU was actually running at under full load tests.
In my own system with to tests temps I use the same software every time to test my CPU temps under full CPU load. The biggest thing that can change my CPU temps is amb temp. So if my amb temp is 21c my CPU may go up to 68c as a max temp after 30 minutes of full load using a CPU torture test software and yet in the same 21c amb temp in a game the CPU only runs at maybe 47c-49c in a game load. Then if amb temp goes up to 28c in the summer months that same test will raise to 75c under full load torture test. In game temps also raise to 54c-56c in the same amb temp of 28c.
So in my example temps go up as expected with amb temp increase to pretty close or right on the money. I use a 280mm EVGA CLC on my CPU and noticed the temps pretty much flatline after about 20 minutes as the CLC reaches it working temp anything after that seems to pretty hold pretty steady in very long game sessions or full load torture tests. My CPU is a old i7 2600K@5.1GHz with controlled voltage set to max out at 1.47 volts with load line set to turbo on my Gigabyte board if I set it to extreme load line the board then tries to over shoot my set max voltage.
I used to use a big air cooler on this CPU form 2011-2018 and was only able to get 5.0GHz before the cooler would get to hot and 5.1GHz was a no go and required more voltage. After switching to the 280mm CLC I was able to lower voltage down to 1.45 volts for 5.0GHz and 1.47 @5.1GHz was finally a go and have good temps. I feel my testing on my own system has shown me that everything is pretty much spot on and is 100% rock solid stable for such an old platform in 2020.
For a moment there, I thought you had a box that could theoretically show cooling performance, from any given heatsink and or cooler based on the TDP specifications of computer processors!
that would've been actually useful.
@@levyroth That could be useful if the TDP numbers actually had anything to do with power consumption.
I worked for a large American semi-conductor company as a software engineer supporting test teams, particularly on the statistical process control and process capability side, and seeing your methodology is pretty impressive. Maybe the best I've seen from an "amateur" setup (meaning one without millions of dollars in testing equipment). People probably don't realize the number of factors for which you have to control to get valid results and the amount of variance (and RMA's) you'll get if you don't properly control those variables. Impressed!
At some point, Steve will find out how to escape the matrix. He has gone down the rabbit hole.
At 1:35; so kinda like the Dunning-Kruger effect of CPU Coolers. Thanks GN for the deep dive, that's why I love you guys.
the science is strong in this one
This is incredibly important and underrated work you're doing. Thank you for your incredible diligence and attention to detail.
No one is cooler than Tech Jesus himself
Even though he may be running 95 degrees C internally
@Ziv Zulander he keeps it longer to increase surface area
He might be running 95C internally, but I would hope no one would place a thermocouple in him to find out...
love your work and you guys have basically been the reviewers that i check other reviews against for a few years now, keep up the good work
"People want to see real world data" Sure, I want to see it validated against real world data, but I'm much more interested in that heater and the numbers it would produce. I'd trust that far more than seeing "real world data"
Lets see, we have Hardware Unboxed for Monitors, Actually Hardcore Overclocking for Motherboards, JayzTwoCents for Water Cooling, and Gamers Nexus for Cases and CPU coolers. CPU and GPU data is done fairly universally well from a lot of the big names (Your a big name to me GN =D). Once you get your crazy setup for Power Supplies and case fans done, that is basically the whole computer.
I am more then happy to continue supporting your guys through the GN Store and Patreon with the amount of passion and integrity that you bring to your work. Thank you GN Team.
Usually when someone says “everyone is lying but us” I’m skeptical.. when GN says it I worry for my wallet 😒
Not gana lie, I really liked seeing and learning about the extensive measures you guys go through to ensure proper and accurate data on such a volatile and difficult thing. I've only fussed a little with doing my own testing and realized there was so many variables, so thank you and good job taking the time to break it down and work out how to control them as best as humanly possible! Keep up the good fight against misinformation!
Paste aging and inconsistency - use a carbonaut pad instead?
After watching your vids for almost a year now, this was the one that made me subscribe. Excellent content and I'm impressed with the effort you've put in to reliably and accurately measure cooler performance. Worthy of a doctorate - well done!
took him 33mins to tell us, "Hey guys look at my new toy!" LOL
If it was bz 4hrs on stream would have elapsed before he blurts out his choices based on use case and cost tier.
I used to really dislike sites that used a static thermal testload for cooler reviews. You have done an excellent job of reversing my opinion... hope to be seeing static load cooler results in the future!
Smashing the like button so hard.
The past week I was looking for an air cooler and went through your videos, web page and more and after that I finally gave up as your numbers are the only ones I would trust for something like this 😅
Looking so much forward to seeing the high end air coolers like NH-D15,, 15S, D14 Dark rock pro 4 etc. battle it out.
For some reason I find your thermal testing interesting and almost soothing, even if I don't need e.g. a case or similar, keep up the good work 😊
Question: why use paste with so much effort, when thermal pad has reusability?
While it can be reused, can it be reused reliably? What happens after it's been clamped and removed 100 times? Is it still just as conductive? What about clamping pressure? Are they consistent with different coolers that clamp differently?
It'd introduce more variables, and most users will be using paste, so the best bet is to just use a known good paste with a known good application method
@@iamdarkyoshi and you're all using thermal compound wrong to begin with.
Paste has a consistently better thermal transfer so differences would be more pronounced and recordable. This is a drastic analogy but it would be like putting a/c in a closed off room and measuring how the temperature changes is the room next to it. There’s probably going to be a difference but it would be a lot easier to measure if there were vents.
Man, THANK YOU for this. This was a joy from start to finish seeing how motivated, focused, and most importantly passionate about tech and accuracy you guys are. I can't wait to see more cooler stuff
After all this Tom's Hardware will still accuse you of lying and just tell us to buy it now.
Gamers Nexus in 2069:
And that is why as a reviewer you need to control what you eat, this can invalidate testing because eating foods with high calorific makes your breathing rate increase and the hot air given out is enough to completely invalidate a cooler's performance. We, at Gamers Nexus, have completely standardised what our technicians eat due to this exact issue. Also, all our technicians are required to maintain a constant 15 breaths per minute from test to test so as to reduce variance in that respect.
*Toms hardware activelity sh*tting themselves at the level of precision the real experts delve into lol
This is one of the reasons why I'm subscribed to this channel. The level of thoroughness is outstanding!
But Linus says that a thermal imaging camera is fine..
Linus is just a salesman, he even has issues doing PC builds he's so bad.
Hats off to all of you at GamersNexus for setting the gold standard for testing and reviews. Nobody else I can think of comes anywhere close to the level of accuracy and legitimacy you guys provide and strive for. Keep up the great work!
Invalid results. You failed to take into account:
1. The current Moon phase
2. Your cat's voltage level
3. Whether you have your hair in front of, or behind your shoulder(s)
4. My feelings
5. The median price of chicken tenders
I expected better Steve.
There's nothing else I can say but BRAVO! You could not have done this and the lion's share of us wouldn't know the difference, but this is so above and beyond that it truly elevates and sets a new standard. What you all provide is an incredible service and you deserve all the accolades you get. Simply stellar work!
I hope we get a good PSU testing as well in this decade.... based on how long/much this refined...
Also, since now thermal-paste testing is awaible as well--- i guess...
Thanks for paying attention to the details and knowing what those details are. Things like knowing that CPUs within the same SKU and run are going to have a +/- 10% difference from chip to chip is why I keep coming back to GN and don't mind spending a bit extra when I need stuff you carry.