HWLabs Nemesis radiators are garbage.

Discussion in 'Water Cooling' started by Zoson, Dec 17, 2016.

  1. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    I made the mistake of buying a few 120 GTS and 240 GTS nemesis radiators without closely comparing them to the older variants.

    Here's a summary of why they're total crap compared to the older ones.

    1. Reduced performance compared to classic GT Stealth due to reduced fin density. GTS Nemesis is LITERALLY HALF the fin density of the classic GTS. It's like HWLabs forgot why people buy their radiators. If I wanted a 16FPI class radiator I could have gone with literally any other manufacturer.
    2. Dimpled screw holes make it IMPOSSIBLE to mount anything but a fan and your case. Even things that are only 1mm thick like dust filters. I cannot comprehend who thought this design decision was good. Instead of having 3-4mm leeway, YOU GET NOTHING. This is honestly a huge fuck you to the customers, and I had to buy 35mm m4 screws from the hardware store and then use washers as spacers to mount anything.
    3. Crappy 'carbon' paintjob is a significant reduction in quality over the previous glossy paint jobs. The paint scratches extremely easily. Don't look at these radiators the wrong way or you'll lose giant flakes of paint.
    4. Over the last 16 years I have NEVER had a radiator leak. I've now had two of these leak RIGHT FROM THE START. No punctured tubes, leaking straight from the damn seams.

    Needless to say, I won't be buying more of these. I got them at PPCS, and left similar notes on the product pages there.

    If anyone has a classic 240mm GT Stealth 32FPI that's in good shape for sale, please send me a PM.
     
  2. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
  3. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004
    Meh? http://www.xtremerigs.net/2016/02/15/hardware-labs-nemesis-280gts-radiator-review/5/

    Seems like a perfectly good radiator to me. Near the tops for low fan speed single fan performance. They have the Nemesis GTX range which is near the tops for push-pull high fan speed. So basically they have all their bases covered....

    That being said, the margins between radiators these days are so small you really shouldn't be concerned with it. Buy based upon price / features and what fits in your build. It seems like you got a couple of bad ones, it happens with all products. I haven't seen large reports of problems with HWLabs, ultimately it comes down to how a company handles customer service when something goes wrong.
     
  4. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    It's a good rad, but if yer buying new the Koolance is the one to get. It out cools the Nem at @1850 by 44watts. The big reason to get the Nemesis is if you run thin and low speed which I'm not sure why one would but if you did, the Nem is the one to get.

    http://www.xtremerigs.net/2015/02/11/radiator-round-2015/5/
     
  5. Chapeau

    Chapeau Gawd

    Messages:
    730
    Joined:
    Jul 17, 2016
    Horse for courses - as Erasmus354 said, the GTS are for low fan speed performance, hence the low FPI.

    I've only tried the EK radiators and HW Nemesis so it's a small sample. But the Nemesis was far better quality, didn't smell as strongly of flux and came out of the box perfectly straight and true. The EK was a mess... the ports weren't straight, side plates warped, terrible quality. I also have a micro 92mm rad from HW which was also perfect. The build quality is awesome..
     
  6. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    thanks for the rec. I'll look into it.

    The HWLabs GTS radiators were *always* the highest density offering they had. The SR series was always the low density offering.


    The problem is that the GTS series was always their high density offering. Now it's not.

    GTS were always for high fan speeds historically. Up until the nemesis series they were 32FPI. And you're flat out wrong about them being for low fan speeds. The SR series from HWlabs are the low fin density rads.
     
  7. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004
    Who cares if they repositioned their product range? They have a low speed radiator that is very good at what it does, and a high speed (in the GTX series) that is also very good at what it does. Instead of complaining that the GTS isn't the GTX just buy the GTX (or not...).

    As for why anyone would want anything other than max fan speed? Some of us water cool for silence and not necessarily for max performance....
     
  8. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    Thin rads and low speed is like cold pizza, to some its still good but its not hot and fresh that's for sure.
     
  9. Chapeau

    Chapeau Gawd

    Messages:
    730
    Joined:
    Jul 17, 2016
    Unless you don't have another option, in which case these are perfect!
    I couldn't find any better for my build...

    There isn't all that much to play with when you're making an apples to apples comparison for slim rads.
     
  10. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    The old GTX was 20FPI also. The new is 16. The GTS was 32, and now is 16. The SR series was 9, and remains 9. The point of all of this is the fin design did not change, and the overall cooling capacity has been reduced vs the last generation. The product line has collapsed and is no longer about the maximum cooling in smallest space with high static pressure fans.

    Again, yes it was my mistake for not looking more closely. This is a PSA that they are not what they used to be.

    If you look at the new koolance that was mentioned above, it's a super dense splitter fin design like the GTS used to be.
     
  11. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    ^^The old GTX at 20FPI, that FPI spec really belied its true nature. On paper it looks like a iddy biddy middle tier pussycat but it is really a P/P high rpm monster.
     
  12. LordGardenGnome

    LordGardenGnome [H]Lite

    Messages:
    78
    Joined:
    May 5, 2005
    You can't really blame a company for your negligence and not paying attention. I just bought two Nemesis GTS 360mm's (push only) for my new build, they're just going to cool my video cards, I went with a Nemesis GTX 280 for my (push and pull) CPU/MOBO/Vreg cooler. If you want SR2 performance look at their new GTR line. I'm building my next water cooling set up on the fact of not having a single fan over 26db of noise.
     
  13. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    The thing is, it makes sense though. Most people do watercooling for better cooling at less noise. A high FPI radiator completely goes against that purpose, and most likely did not sell very well compared to thin low FPI offerings (the GTS was HWLabs only thin radiator, and truly one of the thinnest radiators on the market). Most watercoolers want to run 800-1400 RPM fans, not 2000+ RPM fans, which the original GTS was really geared for.
     
  14. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    Who buys thin rads to go quiet? It's so wasteful. Actually I find that hilarious, go thin and low density. Talk about throwing cooling out the window!
     
  15. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    Did you even actually read my post? SR2 is low FPI, GTS used to be high FPI, and GTS is now mid FPI. None of what you said makes any sense at all in context of the discussion. It looks more like you just wanted to talk about your setup in someone's post regardless of the topic. Thanks for that.

    Missed the point entirely. And also just wrong. You could run quiet fans on a GTS and get *acceptable* performance since they're thin. Which meant you could run a setup of fans that run speed controlled based on temp. They were the ideal radiators for someone that wanted silence at low workloads, but also high cooling capacity at heavy workloads. They already had a low AND mid FPI radiator. Not to mention splitter fin which helps immensely. They completely closed off a market and gave it to koolance.

    Exactly.
     
  16. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    People who couldn't fit a thicker radiator for whatever reason.

    How big is that market exactly? The GTX gen 2 (pre nemesis) outperformed the GTS at fan speeds well into 3000 RPM. The GTS didn't scale all that well until push/pull 2500+ RPM, and if you can fit push/pull, you can fit a larger radiator with one set of fans. A moderate number of fins (15-20) will scale better into 2000 RPM than a 32 FPI design.
     
  17. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    That reason has nothing to do with going thin and low density and nothing to do with your point which was low noise. Not being able to fit a thicker rad in is a logistics problem, not one of design or preference. And at that point it's make do with whatever fits which is so far from yo I like low noise.
     
  18. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    The market with space limited cases but people wanting to watercool for less noise is far greater than the market for people wanting to watercool with slim radiators and high speed (3000+ RPM) fans.
     
  19. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    No one needs to use 3k rpm fans with thin rads. Easy with the hyperbole man, geeze. Go look at the Koolance HX-CU. They don't even test it above 1850rpm lol. I linked the review in post 4. No one reads links anymore or what. I'm not really sure why anyone is arguing, just look at the data.
     
    Zoson likes this.
  20. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    1850 RPM Gentle Typhoons are equivalent to most 2000+ RPM fans on dense radiators. In any case, my 2150 RPM Typhoons are just barely above quiet at 50% (~1000 RPM), let alone "just audible" at 1850 RPM as the article claims. And that 44 watts means what, you're at best 2 C cooler at 2000+ RPM, while being hotter at anything underneath. Again, which market has greater demand? Low RPM slim, or high RPM slim?
     
  21. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    This argument is totally bogus. Anyone who cares about idle temps at low rpm has no idea what they're doing. You can idle at 50c and it will be no detriment to your processor. It's at high load when you're running 100% on 16 threads and you want to stay below 70c that watercooling matters.

    The point, again, is that HWLabs *already* had products to fill the segment you are talking about, all they did was remove the highest performance offering and make it equivalent to their midrange offering. That, is the point. Removal of an entire segment. Removal of choice. Removal of what HWLabs was *all about all along* and discarding the customer base that used them for that reason.

    Also, again, the entire argument that these radiators are crap is less to do with the removal of choice and more about the fact that I had TWO brand new nemesis radiators(one GTX with the wrinkle coat, and one GTS with carbon coat) leaking as soon as they were installed. Literally, day 1, hour 1. The not paying attention to the collapsing of their product offerings was my mistake. Radiators not properly being pressure tested is a HWLabs issue that has nothing to do with me.

    Also let's consider the fact that the koolance radiator mentioned here is clearly a HWLabs original GTS design with splitter fin. HWLabs has licensed the technology to koolance and butchered their own offerings at the expense of their loyal customer base. It's bullshit and poor business.
     
  22. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004
    So I apparently totally missed that HWLabs has three radiators in their lineup now.

    The already mentiond GTS and GTX which indeed have been optimized for low speed low noise performance. They also recently released the GTR series which is optimized for high speed fan performance. http://thermalbench.com/2016/11/22/hardware-labs-black-ice-nemesis-gtr-480-radiator/5/ shows that the GTR performed the best of all radiators tested once the fan speed got near max. (also showed that the GTS and GTX were some of the best at low fan speed).

    An important thing to notice, there has been a lot of focusing in this thread by some on the specific FPI count as if it were truly representative of performance. The GTR and GTX/GTS all have 16 FPI, however they have vastly different performance characteristics. So it is important to not hold preconceptions about something based upon FPI alone, and to remember to look to testing.
     
  23. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    FPI is a rough estimation of density but it is only a physical count, it doesn't quantify split fin designs and other fold designs. And that review smacks of a hwlabs advert.
     
  24. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    You have no idea what you are talking about, and you made that clear the moment you said the GTS was HWLab's top radiator. For that generation, the GTX was the top performer, hence the name GT EXTREME. The GTS is GT STEALTH. Here is a couple of the marketing that HWLabs used for the GTS:

    Which it failed to do compared to the competition. Other slim designs proved to be better performers at quiet levels.

    Compare the marketing to the GTX:

    Fin density do not determine the level of performance. A combination of fin density and thickness does. The GTX takes it one step further with the front to rear flow design rather than the U design, making use of the principal of counter-current heat transfer, a design that is carried into the Nemesis GTX and GTR.
     
  25. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    Actually, NOWHERE did I assert that it was the highest performing overall radiator. The only assertion I made was fin density in the same formfactor was reduced. Which would directly affect performance if you use sufficient fans.

    Sorry, you're arguing with someone who clearly knows more than you do. You literally have no argument and you're trying to make crap up at this point while attempting, and failing, to argue semantics.
     
  26. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    Sorry, come again?
     
  27. Zoson

    Zoson [H]ardness Supreme

    Messages:
    6,264
    Joined:
    Aug 4, 2001
    Nice super fail. I was responding to your complete wrong information that the GTX was 16FPI all along. It was 20FPI previously and has been reduced to 16 along with the GTS. You have nothing. Literally nothing. You're so blatantly wrong that you're grasping at straws making up crap... AGAIN.
     
  28. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    We have been discussing the GTS radiator the majority of the thread. When did you mention in the GTX in the context of the discussion? Also, nowhere did I say the GTX Gen 2 has 16 FPI. It seems like you are the one with communication and comprehension issues, as well as some sort of complex.
     
    Last edited: Jan 6, 2017
  29. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004

    You need to get over your hangup about FPI being all that meaningful. I have already shown testing here that proves that it really isn't. Other factors such as the style of fins, the number and thickness of tubing runs, the thickness of the radiator, and the depth of the plenum play just as much of a role into how well a radiator performs under different fan speeds and conditions. You can't simply cite an FPI number as an example of what situations a radiator is designed for.
     
  30. Napoleon

    Napoleon Gawd

    Messages:
    996
    Joined:
    Jan 27, 2003
    Reviving this thread: I'm looking at the HWlabs Nemesis 280 GTS X-Flow Radiator. Does anyone have actual experience with this radiator? Does only the OP have a HWlabs radiator?

    I don't care how the Nemesis GTwhatver comepares across generations, I'm just after performance numbers. I think the conversation about fin density is moot and reviews , if this is legit, like the following should clear up the performance at different fan speeds: HWLabs Nemesis 280 GTS X-Flow Radiator Review

    OP - Were you able to return/replace the leaky radiators? Obviously that is an intolerable mode of failue

    Thank you for the help. In case you're wondering, I'm hoping to use 2x 280mm GTX X-flow with variable speed fans for a CPU/GPU loop with 1/2 ID tubing (converting existing rad setup)
     
  31. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    They are one of the best, if not best below 1300rpm, they start to taper off above 1500rpm. They do not scale well at high rpm because they are tuned/made for 750rpm-1300rpm. The perf numbers are in that link you posted. Look at the chart 750rpm = 167w dissipation, with two rads that gives you 334w roughly. That means you'll be able to cool a 334w load to 10c delta water temp. If you're using that 5670 you'll be looking at upwards of 300w power draws, then you have your gpu to add to the mix. As you can already tell, 750rpm is not enough to keep up at load, so you'll run 1350rpm ish at load. That gives you 500w of cooling at 10c delta. That should be enough to fit your cpu/gpu just inside that envelope.
     
    TheHobbyist and Napoleon like this.
  32. Napoleon

    Napoleon Gawd

    Messages:
    996
    Joined:
    Jan 27, 2003
    That's what I was hoping for; I imagine newer CPUs would have < TDP than my current load, would you agree?

    They're also a steal of a value at $63/piece at PerformancePCs (althought out of stock today, I wonder if this is a bad sign), ~ $126 for 334/504/626W of cooling. Do any other websites sell/stock these?
     
    Last edited: Mar 7, 2017
  33. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    I dunno where you could buy them other than PPCS. Hmm is frozencpu... last I heard they are open under new management but I haven't been back there since the fiasco.
     
  34. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004
    Keep in mind, you dont need to maintain a 10C delta. That is a good performance target for overclocking and great cooling, but you can easily go much higher on the delta and still cool very well.
     
    TheHobbyist, Tsumi and Napoleon like this.
  35. Tsumi

    Tsumi [H]ardForum Junkie

    Messages:
    12,545
    Joined:
    Mar 18, 2010
    Yep. Go to a 15 C delta and you can dissipate 1.5 times as much heat, or about 460 watts of cooling. It really depends on what your cooling goals are.
     
    Napoleon likes this.
  36. Napoleon

    Napoleon Gawd

    Messages:
    996
    Joined:
    Jan 27, 2003
    Sounds like there's no reason I should go beyond 1300 rpm with that double rad double fan setup. I'm not one that runs my comp 24/7, I run it for a few hours of gaming and shut it off or standby. I may run it a full day but rarely overnight. I just enjoy low temps and high performance while it's on :)
     
  37. thesmokingman

    thesmokingman [H]ardness Supreme

    Messages:
    4,838
    Joined:
    Nov 22, 2008
    Ah no, you want the best delta you can get. You'll see that delta temp does not equal what temp your cpu/gpu will be running at. Fundamentally watercooling cannot remove all the heat, which is why we strive for the best delta possible. 10c delta is good for one block, throw in a second and you will see that delta double, etc. as well as core temps jumping. When talking of temps it is in context of full load, idle temps don't matter because at it will be so good as to not matter in the conversation.
     
  38. Napoleon

    Napoleon Gawd

    Messages:
    996
    Joined:
    Jan 27, 2003
    I assume that the temp delta is how much the water heats up at a given thermal load, correct? Yes this would not factor in the heat transfer between the water and CPU and/or GPU block and I would expect their temps to be always be greater. I am not sure what 4x 140mm fans (not sure what brand to get) even sound like, so perhaps I will run it higher all the time. I just like to have options knowing I can potentially dissipate a lot of heat.

    Currently my delta is likely much better than 10C (haven't measured), but that happens when you have two submerged copper heater cores transferring heat into a blue 40 gallon drum of water :). Not a very portable thermal solution though...
     
  39. Erasmus354

    Erasmus354 [H]ardForum Junkie

    Messages:
    13,477
    Joined:
    Mar 12, 2004
    Delta is the temperature difference between the coolant and the ambient air. Basically, you are within a logical range for your cooling capacity, turn the fan speed higher or lower to fit your personal cooling and noise preferences. I was just pointing out that you dont *need* a 10C delta, that is actually somewhat aggressive and is more of an enthusiast performance indicator.
     
  40. Napoleon

    Napoleon Gawd

    Messages:
    996
    Joined:
    Jan 27, 2003
    Yea, I don't really 'need' any of this haha. Thank you for the help