One thing I'd like to see is to have the coolers' thermal resistance normalized for noise. I understand that's not an easy task, but it would be nice to see how much each can dissipate at a given noise level (the only two metrics most people really care about), rather than having both noise and resistance being variables.
I agree, that would be the best data. If not feasible, I'd like to see the thermal characteristics of each cooler using the same fan at the same speed (tested at many levels).
There are way more variables loli, what kind of case do you have? Any added insulation? What's the average sound level in your room? Do you have a high powered GPU that drowns out the CPU cooler? Do you have loud case fans? Do you have a loud PSU? I'm sure you have a very nice system so having a normalized test would clear up the questions for everyone rather than anecdotal "I can't hear it" statements. Not trying to attack what you are saying but I 3rd the idea that normalized testing for performance at a specific sound level would be great. Even then, different sounds can be more grating than others depending on the pitch. This is probably the hardest area of computing to test, is sound.
Same here, R7 1700 + Thermaltake NiC L32 (It's a Slim 140mm tower - why wasn't in the review?) 500 RPM idle, ~600 RPM load... I can't hear it, even trying. The case has 4 CM Silencio PWM @ 900 RPM, and only the air pushed is audible.
I would prefer to see coolers sped up to the point they produce a fixed amount of noise (say 38db or so) and then compare how much they cool. A much better way to compare them.
Interesting topic but I have a hard time getting useful information out of this article. I would have prefered to see noise levels and efficiency at idle and at full CPU load instead of these artificial 7V and 12V levels. Also there is no easy to read conclusion to help people in a hurry choose the best cooler for their need, and I have no idea what the thermal resistance values mean.
When it comes to air coolers, there's Noctua, then there is everyone else. They are superior in virtually every respect, especially quality and support.
For online comments, there's shills, then there is everyone else. Shills which are just trying to spread around this meme that X company's products are the best thing sliced bread.
Just buy whatever you want, regardless what anyone else thinks. Don't pay brands any particular attention, just buy the best product you can for the best price you can and pocket your savings.
Nice, mature response, with the name calling and all. Grow up, then read the actual review. The reviewer said the same thing I did. Now, go read a few more reviews of coolers, and you'll find that virtually all of them agree with this review. Noctua is expensive, but worth it. Generally I agree with you - it doesn't pay to go by brand name - but there are exceptions, and Noctua is one of them.
Noctua also costs a lot more than the competition. So when you're adding 35% to the price tag of any of the other products tested then you usually expect better performance (not mandatory though). So drawing the conclusion that Noctua is king of the hill in general isn't that useful.
So... there are expensive products, then there are cheaper products :). For some people 35% markup might not be worth the extra performance.
Does Noctua pay you in stock, dividends, or a salary to promote their products online? If not, why do you go out of your way to promote free advertising to another company who you owe nothing to? You made your trade; your money for their product, there's no stipulation that you have to continue to sing praises for their company for other people to be more negotiable to give that same company more money for more products.
Likewise, if you're upset about being called out on literal shilling, then it's really you that needs to grow up.
Additionally, the performance roundup review written by E. Fylladitakis doesn't have anything in common with your post, which you attribute as stating that "(t)he reviewer said the same thing I did". Fylladitakis makes no catch-all branding statement. They evaluate the four products, they make the measurements, they compare the measurements, then make a conclusion about the products based on the measurements.
YOU, however, are making a blanket statement about all performance CPU cooling products developed by Noctua being "superior in virtually every respects, especially quality and support."
Those statements can't be anymore unalike in any way. The former was a controlled experiment to evaluate product performance followed by an analysis of the products, the latter (your comment) is an opinion; void of any metric of data or analysis.
If you look at the thermal performance in this roundup, it shows that the Thermalright cooler (at $47, the cheapest of the four measured in this roundup) to have the best thermal performance at low temperature loads (where most performance PCs, the target audience of these $50 tower cooling solutions, stay outside of gaming or benchmark loads), theorized by Fylladitakis to be due to the direct contact design of the heatpipes, and a close-second (literally 16.6 vs 16.8 degrees Celsius delta over ambient, figures well within a margin of error of measurement) at 150W loads (where most overclocked mainstream CPUs will sit at full load), and still competes closely at completely unrealistic synthetic 340W loads.
So given a $47 price for the Thermalright vs $65 price for the Noctua, given similar performance figures, it's pretty apparent which has the better value for mainstream performance PCs.
Yet, my post gets deleted, but not his, despite its divisive, hateful, argumentative, obtuse, disrespectful, inappropriate nature. Sounds like four letter words aren't the only thing not allowed on Anandtech. Fairness has no place, either.
I have made my ruling. I give you guys exceptional leeway as I believe you need to be free to discuss technical matters, but I draw the line at profanity and pointless ranting. You guys are done, please move on.
If you wish to discuss it in private please email me.
I think the thermalright cooler is pretty hard to ignore unless you are running a high-wattage CPU. Mild overclocks of a typical 80w CPU will make the thermalright the ideal solution at ~100w load, and also the quietest.
The problem with E. Fylladitakis 's style of testing is that they miss major factors in how the things tested actually function in relation to what they are made for.
The reason why Noctua always does "better" in these synthetic tests is that they generally have a 100% flat contact surface.
The problem with 100% flat contact surfaces is that CPU IHS aren't anywhere close to 100% flat surfaces.
Most CPU IHS are concave, which is why Thermalright HSF always have convex contact surfaces.
Clamping pressure is also highly important, especially for Intel's non-soldered IHS CPUs.
Higher Clamping pressure both reduces the distance between the silicon and the IHS as well as forms a better mating between the IHS and the convex Thermalright contact surfaces.
These, and many other major factors that crop up in the real world make the 100% artificial testing like E. Fylladitakis conducts in actually have an extremely large margin of error, making them far less useful than the testing "scienciness" would lead you to believe.
If what you are suggesting is true, we should see Thermalright outperforming Noctua in every other publication's testing, correct? Most review sites (I believe all other sites actually) test CPU cooler performance on an actual CPU running tests, and more often than not inside of an actual computer case. Yet those same tests bring very similar results to what Anandtech has shown, which would appear to invalidate your entire postulation. How do you explain the lack of disparity between these other journalists' "real world" cases, and what Anandtech has done?
I don't know why I bother posting on this shill infested site anyways, waste of my time.
I'm not going to spoon-feed you for 20 posts like the forums.
I've spoonfed you for literally 100 posts before on the forums and your shill self has never acknowledged anything, making this a pointless conversation by any metric.
Have fun shilling with the other shills, adequate journalism in technology died quite a long time ago, and it shows.
Uh, I'm not on the forums, have you been taking mushrooms? I did read that they are the "least dangerous" psychadelics, but you seem to have overindulged. Apparently my point was irrefutable as you chose not to refute anything I wrote.
The testing methods used in these HSF reviews are perfectly adequate because they remove a number of uncontrollable variables that would result from testing with PC hardware. The simulator equipment can produce repetable results with little to no variance between tests within AT's limited budget. I much prefer artifical tests as the basis for relative comparisons since the tests performed by other review sites won't accurately emulate my specific computing environment anyway and are therefore only useful as similarly relative comparisons. The science of these results appear trustworthy.
IMO, it looks like the Thermalright is the the winner. It's the least expensive, and up to 150 watts, it keeps the lowest temperature. How many CPUs pull over 150 watts? Especially in real world workloads, not torture testing??
That would seem to be the case here, yes. Performs as well as the Noctua (better than the Noctua at low loads and nearly equal at high loads), while being physically smaller, and cheaper, too.
The black and silver finish is relatively attractive, too, but appearances are subjective anyways.
That only remains true if the CPU produces power over the same area as the thermal cartridge used in the test. The problem, specifically with Intels latest 4-Cores, is that they generate 100W on a much smaller area. Then the overall cooling capability and the vertical thermal resistance of the cooler become less relevant, and the lateral thermal resistance of the cooler base-plate becomes increasingly critical. That is why an i7-7700K will run much hotter than older CPUs with the same TDP, but much more die area.
It's also partially attributable to Intel's usage of poor TIM between the die and the integrated heatspreader, rather than their older usage of fluxless solder.
Intel's newer chips (due to smaller die sizes and poorer thermal interface for the integrated heatspreader) means that by the time the heat meets the heatsink vs integrated heatspreader contact area, it's not being dissipated as efficiently as older CPU models.
I think you have the situation backwards. The way the Thermalright excels at lower temperatures and it's direct heatpipe design indicate it has superb transfer from the socket.
It only falls behind later when when the smaller heatsink array can't dissipate it as quickly.
Other sites reviewing coolers (eg X-Bit Labs R.I.P.) have found ways to plot cooling ability vs noise level. That makes it much easier to evaluate and choose the best one. IIRC one site even got cost into the same picture. In any case, leaving the reader to separately juggle delta-T and dBA is weak.
Gold standard for standardized testing the efficiency of noise to cooling is Silent PC Review.
Unfortunately, the site suffered a serious lack of ad revenue and seemed to go into the media death spiral of lack of readership>lack of revenue>lack of content>lack of readership. Hasn't been a posting since August.
Well the article measures perf as shipped which while good is not apples to apples as they all use diff fans. They should be measured separately with the same fan installed into each to take fan speed, voltage and noise out of the equation.
I see this same type of comment applied to GPUs and cases too. Products should be tested exactly as they are sold. If a company wants to get their price under $x and they skimp on the case fans to do it, the reader should know that so they can price it appropriately by adding in the cost of good fans. Conversely if a company's price is higher but out of the box has the performance you would achieve from changing out the fans then that should be acknowledged and the reader be made aware. The way Anandtech does it is absolutely appropriate and allows the reader to get an accurate picture of what they are getting for the money.
right, and like i said that is good, but when it comes to evaluating what product is best at specific noise levels the picture becomes less clear.
isolating the performance of each cooler by using the same fan on each would help in that regard, and would be a cogent data point regardless imo.
in your testing methodology page you didnt mention whether you were using the boxed tim, or something like arctic silver for all products, as it is commonly accepted as the standard by which other tims are measured against in tech circles. i prefer mx-4, because its easier to use, but its good to have a standard to measure against.
you might do that to eliminate one variable from your test, and the same argument could be made for using one common fan across the board for that reason.
Some comment on the difficulty of installation would be helpful. The inlcuded Noctua cooler has repuation of being easy to install while the Thermalright is not.
Late reply, but Thermalright's installation method is pretty simple these days. I'd say a half-step below Noctua's mounting system, but you're not going to be struggling to pin down sprung screws.
It would be interesting to see how these heatsinks perform when they all use the same fan. Maybe test them all with the Noctua fan or some other popular variant (Corsair, GentleTyphoon, SanAce, etc.)
The efficieny and temps of Ryzen are so nice that you don't need AIO or big tower coolers, most of the time the stock cooler is more than enough, and it's silent :D
I think the best cooler for Ryzen is the Coolermaster Hyper 212X, everything else is basically overkill.
I wish there was also a passive test. So we can actually see how that heatsinks on their own performance to gauge the efficiency and quality of the designs themselves. The fans add another variable. Maybe even have a case fan for some airflow. I'd like to see how the heatsink itself does because there can be situations where maybe there is a superior heatsink design here but the company has a poor fan compared to another competitor. The competitors fan could just be superior, thus brute forcing the better temps. How would you really tell? There's no passive test.
Hm, why doesn't the article list the weight of these coolers? When they're so large they tend to add quite some torque to the motherboard, so weight should be an important criterion when choosing between these.
Great testing! TRUE Spirit 140 Direct did exceptionally well for an economy cooler. Would be interesting to see how the TRUE Spirit 140 Power, Archon IB-E X2 and some of the other Thermalright coolers. The PH-TC14S definitely is not a great cooler.
The fact you threw in the Wraith cooler makes me sooooo happy! I was hemming and hawing between 1600 and 1600x with the free cooler thrown in being a nice boost for the 1600. Seeing that the better coolers show a pronounced improvement in most cases tells me that free cooler shouldn't be the decider.
Thermalright is like Cooler Master, simple design that reduce cost but performs well. People that have been OCing for the last 20 years know Thermalright's design is usually test proven
I own two of the original (smooth surface) Thermalright True Spirit 140. Its a very good cooler in many respects save for the fan mounting system. They have the most weak, fiddly wire fan mounts I've ever seen. Plus the rubber anti-vibration pads they were using back then simply won't stay on. I decided with later builds to use Noctua in subsequent builds and even though they're about a quarter to a third more expensive its worth it just for their great mounting system. Plus they regularly send me emails to remind me that all my coolers from them are eligible for free mounting hardware for new sockets. No other company that I know of does that.
I think that one important aspect of any cooler review should be the *weight* of the cooler.
Considering that in most system configurations the motherboard is mounted vertically and coolers are hanging off the motherboard, you really don't want them to be too heavy. Even though the cooler may be securely attached to the motherboard the heavier it is the more stress it exerts on the motherboard. Preferably, I don't want mine to weigh much in excess of 500g.
I think that one important aspect of any cooler review should be the *weight* of the cooler.
Considering that in most system configurations the motherboard is mounted vertically and coolers are hanging off the motherboard, you really don't want them to be too heavy. Even though the cooler may be securely attached to the motherboard the heavier it is the more stress it exerts on the motherboard. Preferably, I don't want mine to weigh much in excess of 500g.
Quote - Nevertheless, the thermal performance of the NH-U14S is significantly superior as well, especially when the cooler needs to handle a high thermal load.
Yes, I realize this thread is three years old, however, having come across it I couldn't let that blatantly wrong observation from the review go uncontested when the reviews own results CLEARLY show the Thermalright True Spirit Direct 140 having a lower overall rise over ambient than any of the other three coolers.
How do you figure the Noctua has better performance if your own results show that the Thermalright does? Why did NOBODY notice that, and comment on it?
If Crashman or Garrett over at TH had done that, they'd have been ripped to pieces.
I just read TH’s Best Coolers for 2022 article and the first one in the list is a Cooler Master that fails against the competition in every metric, according to the data in the site’s own reviews.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
74 Comments
Back to Article
Yuriman - Wednesday, May 24, 2017 - link
Nice review! Please do more of these.One thing I'd like to see is to have the coolers' thermal resistance normalized for noise. I understand that's not an easy task, but it would be nice to see how much each can dissipate at a given noise level (the only two metrics most people really care about), rather than having both noise and resistance being variables.
Paapaa125 - Wednesday, May 24, 2017 - link
I agree, that would be the best data. If not feasible, I'd like to see the thermal characteristics of each cooler using the same fan at the same speed (tested at many levels).Lolimaster - Thursday, May 25, 2017 - link
Ryzen 7 1700 + Hyper 212X <45°C at load @900rpm, can't even hear the thing.fanofanand - Thursday, May 25, 2017 - link
There are way more variables loli, what kind of case do you have? Any added insulation? What's the average sound level in your room? Do you have a high powered GPU that drowns out the CPU cooler? Do you have loud case fans? Do you have a loud PSU? I'm sure you have a very nice system so having a normalized test would clear up the questions for everyone rather than anecdotal "I can't hear it" statements. Not trying to attack what you are saying but I 3rd the idea that normalized testing for performance at a specific sound level would be great. Even then, different sounds can be more grating than others depending on the pitch. This is probably the hardest area of computing to test, is sound.Lolimaster - Saturday, May 27, 2017 - link
Cheapo case, only fans are the cpu, psu (a seasonic) and a RX560 a low rpm + undervolt.The only thing I actually hear are my 3 HDD's, if I boot without those is basically the expected electric hum.
My place is quite silent.
Gigaplex - Monday, May 29, 2017 - link
If you can't hear the Hyper 212X @900rpm you may need to get your ears checked.JocPro - Monday, May 29, 2017 - link
Same here, R7 1700 + Thermaltake NiC L32 (It's a Slim 140mm tower - why wasn't in the review?) 500 RPM idle, ~600 RPM load... I can't hear it, even trying. The case has 4 CM Silencio PWM @ 900 RPM, and only the air pushed is audible.jospoortvliet - Friday, May 26, 2017 - link
I would prefer to see coolers sped up to the point they produce a fixed amount of noise (say 38db or so) and then compare how much they cool. A much better way to compare them.Robotire - Wednesday, May 24, 2017 - link
Interesting topic but I have a hard time getting useful information out of this article. I would have prefered to see noise levels and efficiency at idle and at full CPU load instead of these artificial 7V and 12V levels. Also there is no easy to read conclusion to help people in a hurry choose the best cooler for their need, and I have no idea what the thermal resistance values mean.mgilbert - Wednesday, May 24, 2017 - link
When it comes to air coolers, there's Noctua, then there is everyone else. They are superior in virtually every respect, especially quality and support.JoeyJoJo123 - Wednesday, May 24, 2017 - link
For online comments, there's shills, then there is everyone else. Shills which are just trying to spread around this meme that X company's products are the best thing sliced bread.Just buy whatever you want, regardless what anyone else thinks. Don't pay brands any particular attention, just buy the best product you can for the best price you can and pocket your savings.
mgilbert - Wednesday, May 24, 2017 - link
Nice, mature response, with the name calling and all. Grow up, then read the actual review. The reviewer said the same thing I did. Now, go read a few more reviews of coolers, and you'll find that virtually all of them agree with this review. Noctua is expensive, but worth it. Generally I agree with you - it doesn't pay to go by brand name - but there are exceptions, and Noctua is one of them.close - Wednesday, May 24, 2017 - link
Noctua also costs a lot more than the competition. So when you're adding 35% to the price tag of any of the other products tested then you usually expect better performance (not mandatory though). So drawing the conclusion that Noctua is king of the hill in general isn't that useful.So... there are expensive products, then there are cheaper products :). For some people 35% markup might not be worth the extra performance.
WinterCharm - Thursday, May 25, 2017 - link
So? If you want the best you often have to pay more for it.The "best" is the best.
If you want the "best within a price bracket" THEN you can finagle and weigh things against each other.
JoeyJoJo123 - Wednesday, May 24, 2017 - link
Does Noctua pay you in stock, dividends, or a salary to promote their products online? If not, why do you go out of your way to promote free advertising to another company who you owe nothing to? You made your trade; your money for their product, there's no stipulation that you have to continue to sing praises for their company for other people to be more negotiable to give that same company more money for more products.Likewise, if you're upset about being called out on literal shilling, then it's really you that needs to grow up.
Additionally, the performance roundup review written by E. Fylladitakis doesn't have anything in common with your post, which you attribute as stating that "(t)he reviewer said the same thing I did". Fylladitakis makes no catch-all branding statement. They evaluate the four products, they make the measurements, they compare the measurements, then make a conclusion about the products based on the measurements.
YOU, however, are making a blanket statement about all performance CPU cooling products developed by Noctua being "superior in virtually every respects, especially quality and support."
Those statements can't be anymore unalike in any way. The former was a controlled experiment to evaluate product performance followed by an analysis of the products, the latter (your comment) is an opinion; void of any metric of data or analysis.
If you look at the thermal performance in this roundup, it shows that the Thermalright cooler (at $47, the cheapest of the four measured in this roundup) to have the best thermal performance at low temperature loads (where most performance PCs, the target audience of these $50 tower cooling solutions, stay outside of gaming or benchmark loads), theorized by Fylladitakis to be due to the direct contact design of the heatpipes, and a close-second (literally 16.6 vs 16.8 degrees Celsius delta over ambient, figures well within a margin of error of measurement) at 150W loads (where most overclocked mainstream CPUs will sit at full load), and still competes closely at completely unrealistic synthetic 340W loads.
So given a $47 price for the Thermalright vs $65 price for the Noctua, given similar performance figures, it's pretty apparent which has the better value for mainstream performance PCs.
Ryan Smith - Wednesday, May 24, 2017 - link
I love a vigorous discussion, but you guys need to cool it, please. There is no place for profanity here on AnandTech.mgilbert - Wednesday, May 24, 2017 - link
Yet, my post gets deleted, but not his, despite its divisive, hateful, argumentative, obtuse, disrespectful, inappropriate nature. Sounds like four letter words aren't the only thing not allowed on Anandtech. Fairness has no place, either.Ryan Smith - Wednesday, May 24, 2017 - link
I have made my ruling. I give you guys exceptional leeway as I believe you need to be free to discuss technical matters, but I draw the line at profanity and pointless ranting. You guys are done, please move on.If you wish to discuss it in private please email me.
fanofanand - Thursday, May 25, 2017 - link
We love you Ryan! Even when we complain, whine and moan about x, y, or z not being tested the specific way WE would do it, we love you!Ruh-roh, now I"m shilling for Anandtech! Don't worry, my username checks out.
JackNSally - Wednesday, May 24, 2017 - link
I have to agree, Noctua is the best. I do recommend the Hyper 212 for budget builds though.Samus - Wednesday, May 24, 2017 - link
I think the thermalright cooler is pretty hard to ignore unless you are running a high-wattage CPU. Mild overclocks of a typical 80w CPU will make the thermalright the ideal solution at ~100w load, and also the quietest.Communism - Wednesday, May 24, 2017 - link
The problem with E. Fylladitakis 's style of testing is that they miss major factors in how the things tested actually function in relation to what they are made for.The reason why Noctua always does "better" in these synthetic tests is that they generally have a 100% flat contact surface.
The problem with 100% flat contact surfaces is that CPU IHS aren't anywhere close to 100% flat surfaces.
Most CPU IHS are concave, which is why Thermalright HSF always have convex contact surfaces.
Clamping pressure is also highly important, especially for Intel's non-soldered IHS CPUs.
Higher Clamping pressure both reduces the distance between the silicon and the IHS as well as forms a better mating between the IHS and the convex Thermalright contact surfaces.
These, and many other major factors that crop up in the real world make the 100% artificial testing like E. Fylladitakis conducts in actually have an extremely large margin of error, making them far less useful than the testing "scienciness" would lead you to believe.
fanofanand - Thursday, May 25, 2017 - link
If what you are suggesting is true, we should see Thermalright outperforming Noctua in every other publication's testing, correct? Most review sites (I believe all other sites actually) test CPU cooler performance on an actual CPU running tests, and more often than not inside of an actual computer case. Yet those same tests bring very similar results to what Anandtech has shown, which would appear to invalidate your entire postulation. How do you explain the lack of disparity between these other journalists' "real world" cases, and what Anandtech has done?Communism - Thursday, May 25, 2017 - link
Keep up your rhetorical questions.I don't know why I bother posting on this shill infested site anyways, waste of my time.
I'm not going to spoon-feed you for 20 posts like the forums.
I've spoonfed you for literally 100 posts before on the forums and your shill self has never acknowledged anything, making this a pointless conversation by any metric.
Have fun shilling with the other shills, adequate journalism in technology died quite a long time ago, and it shows.
Zetbo - Friday, May 26, 2017 - link
What a loser you are. When the data does not backup your point of view...you call everyone a shill! Thats the way to go! :DCommunism - Friday, May 26, 2017 - link
Took you a whole day to make another account?You really should get your pay docked.
Keep going and you're getting doxed.
fanofanand - Friday, May 26, 2017 - link
Doxx me big boy :)fanofanand - Friday, May 26, 2017 - link
Uh, I'm not on the forums, have you been taking mushrooms? I did read that they are the "least dangerous" psychadelics, but you seem to have overindulged. Apparently my point was irrefutable as you chose not to refute anything I wrote.BrokenCrayons - Friday, May 26, 2017 - link
The testing methods used in these HSF reviews are perfectly adequate because they remove a number of uncontrollable variables that would result from testing with PC hardware. The simulator equipment can produce repetable results with little to no variance between tests within AT's limited budget. I much prefer artifical tests as the basis for relative comparisons since the tests performed by other review sites won't accurately emulate my specific computing environment anyway and are therefore only useful as similarly relative comparisons. The science of these results appear trustworthy.WinterCharm - Thursday, May 25, 2017 - link
Funny how Thermalright performed better than Noctua at low and high fan speeds, then!guidryp - Wednesday, May 24, 2017 - link
IMO, it looks like the Thermalright is the the winner. It's the least expensive, and up to 150 watts, it keeps the lowest temperature. How many CPUs pull over 150 watts? Especially in real world workloads, not torture testing??JoeyJoJo123 - Wednesday, May 24, 2017 - link
That would seem to be the case here, yes. Performs as well as the Noctua (better than the Noctua at low loads and nearly equal at high loads), while being physically smaller, and cheaper, too.The black and silver finish is relatively attractive, too, but appearances are subjective anyways.
A5 - Wednesday, May 24, 2017 - link
Agreed, and I own the Noctua.I've had Thermalright products in the past and they were excellent as well.
ShieTar - Wednesday, May 24, 2017 - link
That only remains true if the CPU produces power over the same area as the thermal cartridge used in the test. The problem, specifically with Intels latest 4-Cores, is that they generate 100W on a much smaller area. Then the overall cooling capability and the vertical thermal resistance of the cooler become less relevant, and the lateral thermal resistance of the cooler base-plate becomes increasingly critical. That is why an i7-7700K will run much hotter than older CPUs with the same TDP, but much more die area.JoeyJoJo123 - Wednesday, May 24, 2017 - link
You're not wrong, but there's more to that story.It's also partially attributable to Intel's usage of poor TIM between the die and the integrated heatspreader, rather than their older usage of fluxless solder.
Intel's newer chips (due to smaller die sizes and poorer thermal interface for the integrated heatspreader) means that by the time the heat meets the heatsink vs integrated heatspreader contact area, it's not being dissipated as efficiently as older CPU models.
guidryp - Wednesday, May 24, 2017 - link
I think you have the situation backwards. The way the Thermalright excels at lower temperatures and it's direct heatpipe design indicate it has superb transfer from the socket.It only falls behind later when when the smaller heatsink array can't dissipate it as quickly.
Eri Hyva - Wednesday, May 24, 2017 - link
Please, add a test with 9 volts.Arbie - Wednesday, May 24, 2017 - link
Other sites reviewing coolers (eg X-Bit Labs R.I.P.) have found ways to plot cooling ability vs noise level. That makes it much easier to evaluate and choose the best one. IIRC one site even got cost into the same picture. In any case, leaving the reader to separately juggle delta-T and dBA is weak.Galcobar - Thursday, May 25, 2017 - link
Gold standard for standardized testing the efficiency of noise to cooling is Silent PC Review.Unfortunately, the site suffered a serious lack of ad revenue and seemed to go into the media death spiral of lack of readership>lack of revenue>lack of content>lack of readership. Hasn't been a posting since August.
snarfbot - Wednesday, May 24, 2017 - link
Well the article measures perf as shipped which while good is not apples to apples as they all use diff fans. They should be measured separately with the same fan installed into each to take fan speed, voltage and noise out of the equation.Ian Cutress - Wednesday, May 24, 2017 - link
Are you measuring the heatsink, or the product? A lot of these companies go to great lengths to 'optimize' their fan design.fanofanand - Thursday, May 25, 2017 - link
I see this same type of comment applied to GPUs and cases too. Products should be tested exactly as they are sold. If a company wants to get their price under $x and they skimp on the case fans to do it, the reader should know that so they can price it appropriately by adding in the cost of good fans. Conversely if a company's price is higher but out of the box has the performance you would achieve from changing out the fans then that should be acknowledged and the reader be made aware. The way Anandtech does it is absolutely appropriate and allows the reader to get an accurate picture of what they are getting for the money.snarfbot - Thursday, May 25, 2017 - link
right, and like i said that is good, but when it comes to evaluating what product is best at specific noise levels the picture becomes less clear.isolating the performance of each cooler by using the same fan on each would help in that regard, and would be a cogent data point regardless imo.
in your testing methodology page you didnt mention whether you were using the boxed tim, or something like arctic silver for all products, as it is commonly accepted as the standard by which other tims are measured against in tech circles. i prefer mx-4, because its easier to use, but its good to have a standard to measure against.
you might do that to eliminate one variable from your test, and the same argument could be made for using one common fan across the board for that reason.
Infy2 - Wednesday, May 24, 2017 - link
Some comment on the difficulty of installation would be helpful. The inlcuded Noctua cooler has repuation of being easy to install while the Thermalright is not.gradoman - Saturday, May 27, 2017 - link
Late reply, but Thermalright's installation method is pretty simple these days. I'd say a half-step below Noctua's mounting system, but you're not going to be struggling to pin down sprung screws.https://youtu.be/EDRNBCH1lRA?t=1760
sheh - Wednesday, May 24, 2017 - link
Would be interesting to know their weights.Sivar - Wednesday, May 24, 2017 - link
Thank you for the useful, thorough review!Daisho11 - Wednesday, May 24, 2017 - link
It would be interesting to see how these heatsinks perform when they all use the same fan. Maybe test them all with the Noctua fan or some other popular variant (Corsair, GentleTyphoon, SanAce, etc.)Lolimaster - Thursday, May 25, 2017 - link
The efficieny and temps of Ryzen are so nice that you don't need AIO or big tower coolers, most of the time the stock cooler is more than enough, and it's silent :DI think the best cooler for Ryzen is the Coolermaster Hyper 212X, everything else is basically overkill.
Lolimaster - Thursday, May 25, 2017 - link
This will probably be nice for Ryzen 9 monsters.CheapSushi - Thursday, May 25, 2017 - link
I wish there was also a passive test. So we can actually see how that heatsinks on their own performance to gauge the efficiency and quality of the designs themselves. The fans add another variable. Maybe even have a case fan for some airflow. I'd like to see how the heatsink itself does because there can be situations where maybe there is a superior heatsink design here but the company has a poor fan compared to another competitor. The competitors fan could just be superior, thus brute forcing the better temps. How would you really tell? There's no passive test.bug77 - Thursday, May 25, 2017 - link
Hm, why doesn't the article list the weight of these coolers? When they're so large they tend to add quite some torque to the motherboard, so weight should be an important criterion when choosing between these.doyll - Thursday, May 25, 2017 - link
Great testing! TRUE Spirit 140 Direct did exceptionally well for an economy cooler. Would be interesting to see how the TRUE Spirit 140 Power, Archon IB-E X2 and some of the other Thermalright coolers. The PH-TC14S definitely is not a great cooler.fanofanand - Thursday, May 25, 2017 - link
The fact you threw in the Wraith cooler makes me sooooo happy! I was hemming and hawing between 1600 and 1600x with the free cooler thrown in being a nice boost for the 1600. Seeing that the better coolers show a pronounced improvement in most cases tells me that free cooler shouldn't be the decider.Peichen - Thursday, May 25, 2017 - link
Thermalright is like Cooler Master, simple design that reduce cost but performs well. People that have been OCing for the last 20 years know Thermalright's design is usually test provenLeyawiin - Thursday, May 25, 2017 - link
I own two of the original (smooth surface) Thermalright True Spirit 140. Its a very good cooler in many respects save for the fan mounting system. They have the most weak, fiddly wire fan mounts I've ever seen. Plus the rubber anti-vibration pads they were using back then simply won't stay on. I decided with later builds to use Noctua in subsequent builds and even though they're about a quarter to a third more expensive its worth it just for their great mounting system. Plus they regularly send me emails to remind me that all my coolers from them are eligible for free mounting hardware for new sockets. No other company that I know of does that.azrael- - Monday, May 29, 2017 - link
I think that one important aspect of any cooler review should be the *weight* of the cooler.Considering that in most system configurations the motherboard is mounted vertically and coolers are hanging off the motherboard, you really don't want them to be too heavy. Even though the cooler may be securely attached to the motherboard the heavier it is the more stress it exerts on the motherboard. Preferably, I don't want mine to weigh much in excess of 500g.
azrael- - Monday, May 29, 2017 - link
I think that one important aspect of any cooler review should be the *weight* of the cooler.Considering that in most system configurations the motherboard is mounted vertically and coolers are hanging off the motherboard, you really don't want them to be too heavy. Even though the cooler may be securely attached to the motherboard the heavier it is the more stress it exerts on the motherboard. Preferably, I don't want mine to weigh much in excess of 500g.
darkbreeze - Friday, February 21, 2020 - link
Quote - Nevertheless, the thermal performance of the NH-U14S is significantly superior as well, especially when the cooler needs to handle a high thermal load.Yes, I realize this thread is three years old, however, having come across it I couldn't let that blatantly wrong observation from the review go uncontested when the reviews own results CLEARLY show the Thermalright True Spirit Direct 140 having a lower overall rise over ambient than any of the other three coolers.
How do you figure the Noctua has better performance if your own results show that the Thermalright does? Why did NOBODY notice that, and comment on it?
If Crashman or Garrett over at TH had done that, they'd have been ripped to pieces.
Oxford Guy - Tuesday, January 4, 2022 - link
I just read TH’s Best Coolers for 2022 article and the first one in the list is a Cooler Master that fails against the competition in every metric, according to the data in the site’s own reviews.