Gamers Nexus tests Noctua fan rumors.

[ravenshrike]

Finds no significant difference apart from minor color differences. As I suspected, much ado about nothing.

 

http://www.gamersnexus.net/guides/3012-noctua-fans-from-china-and-taiwan-tested

[jcw150]

Thanks, now I don't to watch it 

[iRileyx]

A bit of backstory would be nice.

 

What's he testing them for in the first place? whats apparantly different between (im assuming from the thumbnail) chinese made and taiwanese made noctua fans?

[ravenshrike]

4 minutes ago, RKRiley said:

A bit of backstory would be nice.

 

What's he testing them for in the first place?

A bit ago someone noticed that his fans made in China were a slightly darker poop beige than his other fans and that the stepped inlets had minor shape differences and possibly a 1 dB sound difference. Freaked out about it and started a topic on Reddit. Everybody and their mother started bitching about the chinese fans being clearly inferior quality blah blah blah with only a few voices of sanity. The LTT thread on the subject.

[TrigrH]

TDLR:

China has a new mould for their factory and is the current manufacturer for the fans.

The taiwanese fans use an older mould and are no longer being produced. (old stock)

 

They peform the same. (according to gamersnexus testing)

 

As for the color differences, noctua is looking into it.

[Misanthrope]

Yeah I saw: he had the balls to call 8 fucking fans a "good sample" when it's statistically fucking insignificant.

 

I am really starting to dislike his pseudo-science pretenses here: Much like Salazar (at one point, he's gotten better lately) this guy over-estimates his testing methodology and I have to call into question his results specially when he dismisses things like the AMD Wraith Stealth cooler and casually goes back on that on a future video but without acknowledging he was well, full of fucking shit to begin with.

[Heesleemer]

1 minute ago, Misanthrope said:

Yeah I saw: he had the balls to call 8 fucking fans a "good sample" when it's statistically fucking insignificant.

 

I am really starting to dislike his pseudo-science pretenses here: Much like Salazar (at one point, he's gotten better lately) this guy over-estimates his testing methodology and I have to call into question his results specially when he dismisses things like the AMD Wraith Stealth cooler and casually goes back on that on a future video but without acknowledging he was well, full of fucking shit to begin with.

eh with the amount of fans that would need to be tested, a good sample size would be insanely expensive. with that said I'm more inclined to side with GN and Noctua - as I anecdotally (lol) have seen no real difference in my noctuas (about 10-15 varying from Taiwan and China) 

[Misanthrope]

19 minutes ago, Heesleemer said:

eh with the amount of fans that would need to be tested, a good sample size would be insanely expensive. with that said I'm more inclined to side with GN and Noctua - as I anecdotally (lol) have seen no real difference in my noctuas (about 10-15 varying from Taiwan and China) 

The scope is beyond what I would consider reasonable for an independent tester yes, but that doesn't means that you should call it significant. That's a blatant misrepresentation. He is trying to posse as an authority on testing methodoloy when it suits him but other times he completely ignores his laughable sample size (hint: When it's time to shit on AMD he's all about methodology, when it's time to defend Noctua however, it's "good") 

 

Example: There's a smaller youtuber OztalksHW that I watch and just today he posted a video where he attempted to test i3 vs pentium vs Ryzen 3 in multitasking environments. He specifically mentions how the GamersNexus guy went out of his way to basically tell him that a specific test of having Chrome is inconsistent and would produce meaningless results because there's inherent inconsistencies in the way the load balances on a browser due to caching and other things.

 

Here's the video:

 

Spoiler

 

 

Is this fair advice? Perhaps. The variables are though to narrow down but it seems like he was too technical since using a browser while gaming is a fairly common task users have interest in and the variance wouldn't be that dramatic but I said ok he made his point about inconsistency in testing and the importance of proper methodology to control as many factors as possible, that's fair.

 

Then hours later he releases this piece of shit video "testing" just 8 fucking fans and even calling it "good sample" when nothing can be further from the truth. On it's own I could say "Ok he's just a reviewer I wouldn't expect him to get 1000 or 2000 fans I'd be prohibitive and mostly untestable for a small team within a reasonable timeframe" But putting both together, do you see the fucking inconsistencies? 

 

He's ok with calling out a smaller channel publicly about their testing methodology while at the same time testing with 8 fucking fans only. How full of yourself do you have to be to intentionally interfere with your peers while you release pointless fucking testing that's worst yourself.

[ravenshrike]

23 minutes ago, Misanthrope said:

Yeah I saw: he had the balls to call 8 fucking fans a "good sample" when it's statistically fucking insignificant.

Given the variance, or lack thereof, shown with a single fan reaching at max one dB higher than the others at only one point in its curve, the sample size was perfectly respectable. Does it completely rule out significant manufacturing differences, no, but it makes them much less likely, especially as one of the fans was the purportedly 'problematic' fan reported by the reddit user. Whose testbench involved placing fans on wooden furniture with a piece of cardboard/cloth in between them and one fan less than 6 inches from the wall while the other fan appeared to have a bunch of furniture and cloth next to it.

[Misanthrope]

1 minute ago, ravenshrike said:

Given the variance, or lack thereof, shown with a single fan reaching at max one dB higher than the others at only one point in its curve, the sample size was perfectly respectable. Does it completely rule out significant manufacturing differences, no, but it makes them much less likely, especially as one of the fans was the purportedly 'problematic' fan reported by the reddit user. Whose testbench involved placing fans on wooden furniture with a piece of cardboard/cloth in between them and one fan less than 6 inches from the wall while the other fan appeared to have a bunch of furniture and clothe next to it.

Read above: It's not only about this case specifically but about his extreme hypocrisy in how he presents his stuff and how he complains about what other channels are doing.

[Derangel]

17 minutes ago, Misanthrope said:

Yeah I saw: he had the balls to call 8 fucking fans a "good sample" when it's statistically fucking insignificant.

 

I am really starting to dislike his pseudo-science pretenses here: Much like Salazar (at one point, he's gotten better lately) this guy over-estimates his testing methodology and I have to call into question his results specially when he dismisses things like the AMD Wraith Stealth cooler and casually goes back on that on a future video but without acknowledging he was well, full of fucking shit to begin with.

 

Considering Noctua stated that the NF-A14 was primarily made in China for years I'd argue that, even with the small sample size, Steve's conclusions are most likely 100% correct, outside of some fans that will inevitably fail or not work as well. I agree that he shouldn't say only eight is enough to make a full conclusion, but barring evidence to the contrary I have no problem believing the results. If the switch to China made a difference in the fans we would have heard about it a few years ago.

[ravenshrike]

1 minute ago, Misanthrope said:

Read above: It's not only about this case specifically but about his extreme hypocrisy in how he presents his stuff and how he complains about what other channels are doing.

Except he explicitly states in the video that this is not completely definitive and that if anyone finds fans with issues invites them to report their findings etc... etc... All that's really being done is a refutation of the claim by the reddit user, which as it involved one of the fans GN tested does mean they have a significant enough sample size.

[jcw150]

4 minutes ago, Misanthrope said:

Read above: It's not only about this case specifically but about his extreme hypocrisy in how he presents his stuff and how he complains about what other channels are doing.

I don't have a problem the sample size here but I agree with your point regarding GamersNexus, he seems to live up his own arse.

[Misanthrope]

4 minutes ago, ravenshrike said:

Except he explicitly states in the video that this is not completely definitive and that if anyone finds fans with issues invites them to report their findings etc... etc... All that's really being done is a refutation of the claim by the reddit user, which as it involved one of the fans GN tested does mean they have a significant enough sample size.

Yeah I'm not impressed by a "cover my ass" statement at the very end. Specially given the circumstances mentioned.

[jcw150]

Gamers Nexus needs to learn to do concise coverage or be more interesting

[NvidiaIntelAMDLoveTriangle]

33 minutes ago, jcw150 said:

Linustechtips needs to learn to do concise coverage or be more interesting

Sorry, had to.

[mr moose]

Actually, I believe the sample size statistically speaking is 2.  Because 4 fans in a row from one production line should be almost identical and the 4 fans from another should be almost identical also.  To properly test production line quality the samples would have to be random from different production runs from each country. 

 

It's like testing a lake for algae by getting all your samples from one corner at the same time and saying this lake has less algae than that lake.

[Misanthrope]

13 minutes ago, NvidiaIntelAMDLoveTriangle said:

Sorry, had to.

Nah I'm fine with Linus focusing on precariously juggling expensive hardware: see today's video. 

[rn8686]

Not really supriesed, as I said, just bad moulds. 

[Masada02]

While I generally don't take issue with GN testing methodology, I did take issue with him dismissing the one outlier fan (that was running 1dB louder than the others) from his samples. Sure, it would be fine to dismiss that outlier if it was a massive sample size, but he tested 8 fans and 12.5% of them was noticeably different. That is statistically significant and should have been noted.

[AnonymousGuy]

Someone needs to ask Noctua then what the fuck the purpose of all their fins and grooves and shit are then if they can be fucked up badly and really not make much of a difference.

 

Also, he really should have done a frequency spectrum.  Because I suspect ^ probably translates to reducing noise at specific frequencies that may not contribute much to total dB.

[spartaman64]

glad to know that noctua fans can still be my number one recommendation to someone looking for a premium fan

[Bcat00]

2 hours ago, Misanthrope said:

The scope is beyond what I would consider reasonable for an independent tester yes, but that doesn't means that you should call it significant. That's a blatant misrepresentation. He is trying to posse as an authority on testing methodoloy when it suits him but other times he completely ignores his laughable sample size (hint: When it's time to shit on AMD he's all about methodology, when it's time to defend Noctua however, it's "good") 

 

Example: There's a smaller youtuber OztalksHW that I watch and just today he posted a video where he attempted to test i3 vs pentium vs Ryzen 3 in multitasking environments. He specifically mentions how the GamersNexus guy went out of his way to basically tell him that a specific test of having Chrome is inconsistent and would produce meaningless results because there's inherent inconsistencies in the way the load balances on a browser due to caching and other things.

 

Here's the video:

 

  Reveal hidden contents

 

 

Is this fair advice? Perhaps. The variables are though to narrow down but it seems like he was too technical since using a browser while gaming is a fairly common task users have interest in and the variance wouldn't be that dramatic but I said ok he made his point about inconsistency in testing and the importance of proper methodology to control as many factors as possible, that's fair.

 

Then hours later he releases this piece of shit video "testing" just 8 fucking fans and even calling it "good sample" when nothing can be further from the truth. On it's own I could say "Ok he's just a reviewer I wouldn't expect him to get 1000 or 2000 fans I'd be prohibitive and mostly untestable for a small team within a reasonable timeframe" But putting both together, do you see the fucking inconsistencies? 

 

He's ok with calling out a smaller channel publicly about their testing methodology while at the same time testing with 8 fucking fans only. How full of yourself do you have to be to intentionally interfere with your peers while you release pointless fucking testing that's worst yourself.

8 fans is actually a decent sample size for a single reviewer.  No one is going to buy a hundred fans and test them all, its unrealistic and for the love of god, come back it reality and think twice.

[Skanky Sylveon]

Glad to know that there is practicality no performance difference.

 

The one Chinese fan that was on average 1db louder is a little concerning though.

[tom_w141]

5 hours ago, jcw150 said:

Gamers Nexus needs to learn to do concise coverage or be more interesting

No. Longer more in depth videos = good. If everyone becomes LTT with 6 minute product reviews we will never learn anything meaningful. You can't be thorough and quick. I'd rather a 20 minute review to a 5 minute review. I watch LTT for entertainment but not really for serious reviews/knowledge.