Groove_C

No need to pay for NZXT Kraken or Corsair AiOs that cost much more for higher CPU and VRM temps, billion cables and crap software. - No RGB bling bling - No software - only 1 cable to connect to mainboard PWM fan connector to power the pump and 2x 140 mm fans - 2x 140 mm fans don't need to be connected to 2x mainboard fan connectors - 1x dedicated functional (no gimmick) small and quiet VRM cooling fan - fat 38 mm radiator vs. competition 28-30 mm rads - higher radiator fins density vs. competition - very good and well known thermal paste (MX-4) - very good stock fans (low noise and relatively high pressure) - less than $100 (280 mm model)

@oldbear I have 0 problems with my GTX 1070 Ti Strix OC Advanced Binned working at 2088 MHz core and 9600 MHz memory. Never more than 60°C in any graphics benchmark and it's not loud at all with tripple slot radiator and 3x 92 mm fans. Asus ROG Strix GPUs have the best radiator and fans you can get. It's like Noctua, but for GPUs. EVGA, MSI, Gigabyte and Zotac should take lessons from Asus. EVGA and MSI have not that bad radiators. A lot of plastic on Zotac, Gigabyte and especially on MSI GPUs and not that good fans - not that much static pressure (vs. Asus ROG Strix) and also fans blades are just for looks, which results in not that pleasant sound signature. Gigabyte GPUs have one of worst radiators among current gen nVidia GPUs. Really cheap/poor design and function. Same for GTX 10X0 (Ti) GPUs from them.

Then the only good setup for @cb65 would be all front in and top + rear out.

By the way, 11th gen Intel CPUs will finally have new architecture, even if they will still be in 14 nanometers. They still have problems with 10 nm process. 10th and 11th gen Intel CPUs might be released same year, since it was also the case in 2017, when i-7700K and later i7-8700K were released.

@Chris Ogden I have same CPU and also 32 GB 2400 RAM. For ArmA, your upgrade will bring you maybe 10 FPS, average, not minimal and it's not even multiplayer, but only singleplayer. So no need for upgrade, yet, considering what you have now. Especially considering the cost vs. performance gains. Also Intel will release 10th gen CPUs this or next month. But even then, it won't help you much in A3. With your setup, you can easily wait until something noticeably better, like AMD Ryzen 4xxx at the end of this year or Intel 11th gen CPUs, also end of this year or beginning of next year. So at least, money you will spend, will be better justified, considering higher performance gains for same cost vs. now. Your RAM, if it's DDR3 2400 MHz 10-12-12, then it's on same level as DDR4, somewhere between 3200 MHz CL14 and 3600 MHz CL15. So you're PC is still very good, for A3. It's just Arma itself that's not so well optimized.

@Chimere as said before, I have 2x bottom fans blowing fresh air onto my GPU and I have removed all the dust filters, since I have overpressure in my case with 5 fans in and 3 out. All same model and speed. Have 0 problems with dust. VRMs, CPU, RAM and GPU are all happy.

@cb65 there is no problem using rear fan as intake, since you simply can mounted the fans of your CPU cooler the opposite way, so they both blow air to the front. You don't need 3 top fans, since the 3rd (top front) would blow onto nothing. 1x rear in 2x top in 2x front out

@cb65 if you could specify your case model...

@cb65 This is how the airflow is oriented in my Corsair Carbide Air 740 case. 2x140 mm bottom fans blowing fresh air directly onto 3x92 mm GPU fans 1x140 mm rear fan blwoing fresh air directly onto vertically placed mainboard VRMs and onto the air cooler 2x140 mm top fans blowing fresh air directly onto horizontally placed mainboard VRMs, onto the air cooler and onto RAM sticks 3x140 mm front fans blowing hot air out from the case Have almost no dust inside the case, because of the overpressure, since more air is sucked in than sucked out

Here is a video that reflects perfectly the current situation air vs. AiOs and even custom water. Watch and listen till the end of the video, please. Only 4°C difference for Noctua NH-D15 vs. custom 360 mm water for like 300€ or more. Pay attention to how quietly runs the custom water cooling, thanks to the newest Noctua 120 mm fans. Updated Noctua cooler that will come with more cooling area will not only reduce this 3-4°C gap vs. AiOs or custom water, but most importantly will finally come stock with these updated Noctua fans that are whisper quiet. I think we will see the middle 140 mm fan replaced by the updated version and the existing latest Noctua 120 mm fan in front, instead of another 140 mm fan, like it's the case now on the NH-D15. So we will finally have whisper quiet air cooler and virtually same temp as a custom 360 mm water cooler.

@oldbearthat was also partially my point. Water cooling can significantly decreas3 temps and also further lower noise level, but for this to happen, it must be custom water cooling that can like 700€ and more, for like 200 MHz. The total cost overweights temp, noise and performance gains and you still have failure risk and need of surveillance and maintenance. Custom loops for like 300-400€ result in ~ same temps as AiOs. AiOs... not even need to talk about them. Only aesthetic side, for a price premium, but still failure risk, even if not that high. Possible failure risk results in need to always additionally have an air cooler (another added cost) or if you don't have it, if your AiO fails, you will have to RMA it and wait a relatively long time for replacement, and without possibility to use your PC. But I doubt somebody would stay completely without PC for more than 3 days and will be forced to order an air cooler anyways. So why not order an air cooler directly, instead of an AiO, save money buying air and additional money and bad mood resulting from possible AiO failure and only 3°C difference, but 0 worries, not only for years to come, but for further PCs to come.

With this video, Gamers Nexus (Steve) obliterated all the other youtubers and even all the sites and forums. Before this latest video air vs. water, there was a 30 mins video only explaning how he tests + pages and pages of text and graphs on the site. 99% of other "tests" have flaws, starting from inconsistant thermal paste application and finishing with where, how and with what temps are measured and which conditions and after how much time, like software measurments only, not accounting for the fact that different boards manufacturers and even different boards models from same manufacturers have different brands of sensors of different quality and precision, measuring temp and/or consumption not in same spots on different boards and boards models from same manufacturer + not accounting for current leakage + not using same versions of bios, windows, drivers, bios settings, not even always same CPU model unit, even if it's still same CPU model (same CPU model, but different unit can require more/less voltage for same frequency, which will result in consumption variation between 2 same model CPUs) + not using most up to date CPUs to test newer coolers, but still use like 4th gen Intel CPUs with only 8 threads in 22 nm All of this because most of them are just to entertain the average user and also because they don't want to work hard if people still eat all of that anyways. Otherwise it would require them all to research much more to better understand all of the small but important details, that make all the difference + retest a lot of coolers each time if one or several variables change. And they don't want that, since it costs time and time is money. So better do alomst every day short and "useless" vids that don't require as much knowledge and time and still bring a lot of viewers and money.

If you have any hard facts/proofs that prove I'm wrong, let me see them then. I always backup my points with hard evidence. You always only talk and never backup your words with anything. And like said before, I don't have my own opinion on things, where there can't be personal opinion, where only hard numbers speak and can't be interpreted in any other manner by anybody at all. Temps, noise, reliability can't be an opinion, whatever you say.

So why pay 60€ more for Corsair or NZXT 280 mm AiO vs. Noctua NH-D15 for only 3°C difference at same noise level + possible failure any time, if you don't find air coolers ugly/bulky and they let you plug in and remove your fans cables and RAM sticks of any hight no problems, without removing the cooler? The only valid point that remains -> looks, at a price. Other imaginary points, like lower noise and better temps of your current Corsair AiO and less stress for the mainboard PCB due to air coolers weight are simply not valid at all. Mainboards can cope with such weight no problems. The video proves that to have slightly better temp, AiOs have to run at least at same or higher noise levels than Noctua air cooler. The video proves that further lowering AiO fans speed, will make them only marginally more silent than Noctua air cooler + then there will be 0°C difference for 60€ more + failure possibility. The video proves that 6 years old Noctua fans of same size can have same noise level at much higher speed than newest fans of other brands, that have to be set to much lower speed to keep same noise level as Noctua fans, which reduces their static air pressure and reduces transported air volume. The only thing that saves these overpriced bling bling RGB AiOs is simply having greater coolant volume inside vs. Noctua NH-D15. Updated Noctua cooler that will be released this year, will have greater cooling area and will come with even better and larger fans, while keeping same size. So bye bye 3°C difference and associated/possible failure/risk of AiOs. Even if the updated Noctua cooler will cost 10-20€ more than the current NH-D15, it will still be 50-40€ less than same performance/noise AiOs.

Like I said before and this most up to date video confirms it, temps difference of a 6 years old air cooler Noctua NH-D15 vs. 360/280 mm newest AiOs is only 3°C, on by now "common" 16 threads CPUs with OC for only $90 and 0 possible failure points. Most people that buy high threads count CPUs won't buy an air cooler anyways and even if they were to do so, max temp difference would be 7°C, which is not dramatic at all. But most people still have 8 threads Intel i7. Other people currently buy only 6 cores Intel i5-9600K, only 8 cores i7-9700K or "only" 12 threads AMD Ryzen 5 3600(X).

Everyone, please find time and patience to watch and listen to this video carefully till the end. Released today! It's the most accurate one from all existing vids and forums' discussions out there. Here the guy tests best available current air and water coolers with different (stock) fans and different radiator size, on AMD and Intel. On CPUs with lower clock and voltage but larger contact area and on CPUs with higher clocks and voltage, but smaller contact area. He tests all coolers adjusted to same noise level, to see temp difference in same conditions + He tests all coolers at max fans speed for each, to see temp difference, noise level aside. One can clearly see, that Noctua fans on the NH-D15, that's from 2014 can rotate at higher speed than current fans of other brands and still produce same noise (35 dB), which is the indicator of still best noise/pressure/air volume fans, despite 6 years existance. Current Noctua fans not included in the test, since not available in 140 mm to come stock on the NH-D15 cooler. Tested on 32 threads AMD R9 3950X at only 4.2 GHz 1.219 V 200 W load: At 35 dB (same noise level for all coolers), NH-D15 is 4°C behind best 280/360 mm AiO coolers and this is how it should be tested -> apples to apples -> current hardware in same conditions, side by side. At 35 dB (same noise level for both AiOs), 280 mm AiO (from same brand and model + same pump) is 0.5 °C cooler than 360 mm variant. Tested on Intel HEDT class CPU (not mainstream) 150 W load: At 40 dB (same noise level for all coolers), NH-D15 is 7°C behind 360 mm Corsair H150i AiO. At 40 dB (same noise level for all coolers), NH-D15 is 6°C behind 280 mm Corsair H115i Platinum AiO. At 40 dB (same noise level for all coolers), NH-D15 is 6°C behind 280 mm NZXT Kraken X63/X62 AiO. Same goes for updated X73/Z73 and X63/Z63 NXZT AiOs. At 40 dB (same noise level for all coolers), NH-D15 is 6°C behind 280 mm Corsair H115i v2 AiO. At 40 dB (same noise level for all coolers), NH-D15 is 2°C behind 240 mm Corsair H100i Pro AiO. Tested on 32 threads AMD R9 3950X at only 4.2 GHz 1.219 V 200 W load: At max fans speed for all coolers, NH-D15 is 4°C behind 360/280 mm NZXT Kraken X72/X62. Same goes for updated X73/Z73 and X63/Z63 NXZT AiOs. Tested on 16 threads AMD R7 3800X at 4.3 GHz 1.287 V 125 W load: At 35 dB (same noise level for all coolers), NH-D15 is 3°C behind 360/280 mm NZXT Kraken X72/X62. Same goes for updated X73/Z73 and X63/Z63 NXZT AiOs.

Yes, IceGiant ProSiphon Elite will be pretty heavy, but since it's not a problem for any modern motherboard PCB, there is nothing to worry about and thus no arguments for an AiO, justifying it by moving weight from the motherboard to the PC case. Also as it will be oriented in the same way as a GPU and it will be pretty high, there won't be any problems to plug in any fans or install/remove literally any RAM. What you fail to realize is that I don't try to always have the last word, to be right even if I'm not or convince somebody that AiOs are bad. I can't know everything and I'm always open to constructive critics/feedback and new/different sources/opinions, but only if the opponent refers to solid data one can rely on, so I can simply check it myself as well and take updated/different info from there to have a updated/wider point of view as well. Words only don't matter, like at all. There are things, where there can't be my or your opinion, but only hard numbers and graphs, which is what needs to be provided when discussing or proving your personal observations/experience. One can say that one woman is more beautiful than the other or that one cooler looks nicer/cleaner/slimmer, since everybody has different preferences. Don't think that everybody likes RGB and that everybody absolutely likes and needs a clean view of the mainboard and RAM. I have no problems with brown Noctua color and I don't look at my mainboard and RAM at all. I see no point in doing so, despite having my case on the table and with a glass side pannel. So here you already have 1 point of view that's different from yours. But when it comes to reliability, noise, temps, ease of service or lack of it, only facts have weight, proven by provided numbers and not only by words without any data/proofs. I never said that one should avoid AiOs at all cost and that they're simply bad/dangerous. If you read my messages again, you will (hopefully) see, that I only said, that one should provide data/proofs when trying to prove own point of view/experince and that AiOs simply have much more points of failure than an air cooler that simply has none, even if good quality AiOs have a very very low failure rate and when there are some failures, you can be sure this "news" will quickly spread across the net and this way will strongly distort the real situation. But nontheless, you never know if/when it will fail + considerably higher price than the best air cooler, if you want the most reliable and quite pump and fans, for almost negligible temp difference. If one cares about looks and absolute best possible temp (price/potential risk aside), even with same or lower noise than the most expensive air cooler, then go for it. Once again, it might work flawlessly like 5 years right from the start, but there will always be a chance something can go wrong, best if right from the start, but can occure at any moment over time and you can't do anything about it. I might also switch to water later, if IceGiant ProSiphon Elite won't be good. But I will make sure to have a good pump, with at least 6 poles (better 8 poles) motor, good fans, possibility to replace/disconnect everything for maintenance and with possibility to refil/top it out. NZXT has the most reliable and silent/stable pump of all AiO pumps at the moment (8 poles motor) in their Kraken X/Z 63/73 models. But you can't disconnect hoses from the CPU base plate and from the radiator to shorten them or to replace them by longer ones, better translucent, to see if there is air or sediment/algea build up.

Problems with AiOs are not common, but still very possible, as you can see. 3x RMA for one and same person. Air coolers simply can't have any problems, since no pump to fail, no hoses to come loose, no possibility of sediment or algea build up, despite biocides and no galvanic corrosion (aluminum in reaction with copper), despite special additives. Well it's a well known problem, depending on how many poles pump motor has and if the axle is ceramic or steel. Even regular fans can exhibit noise and vibrations, at very low rpms right from the start or can develope such problems over time, depending on their price. It's "solved" by software not allowing them to run lower than N rpms to avoid this. Like you admit it, it's been quite some time ago you've owned an air cooler last time. 1. You simply can't remember it. And even if you do -> see points below. 2. You don't know/mention what was the speed of the fans on your old air cooler, which you also could have additionally adjusted in the BIOS to be dead silent. 3. You don't mention the speed of the fans of your current AiO as well. 4. One should always compare side by side, to not rely solely on memory and also at same rpms for both, you current AiO fans and those of your old air cooler. This way you could say, that your old Noctua fans at like 1200 rpms make more noise than new Corsair fans at same 1200 rpms. And even then, you're comparing old tech to newest tech, which is ok, but doesn't reflect the current state of air vs. water. Just FYI, Noctua still has the best fans (lowest noise at same rpms and highest air pressure at same rpms), if you compare latest Noctua fans latest fans of other brands, no metter whether used on air cooler radiator, on water cooler radiator or in PC case. Corsair fans are good as well, but not nearly as good as Noctua. Price aside. Once again, telling us, that there is greater dust build up resulting from an air cooler vs. AiO, without mentioning fans rpms for both makes no sense. Also dust filters can be different on your case and your daughter's case and case palcement also impacts temps and dust build up. What I can tell your for sure is that if you take same fans, at same rpms, they will make more noise on an AiO radiator than on an air cooler radiator, since there is more resistance to the air, due to higher number of fins tightly spaced and for same reason - dust will build up quicker as well. Well, relying only on your own experience is not correct, since there are many other users that have the opposite experience to yours, even with same brand you currently have, namely Corsair. You can simply look on amazon and see that there are much more bad reviews for Corsair AiOs than on Corsair's forum. And it concerns any AiO in general, not only Corsair ones, since there are more parts and moving parts etc, unlike in air coolers.

@Chimere my NH-D15S isn't touching anything because of cutouts 😏 And there is even some space left 🙂

@Chimere In the video they were talking about 30 mm radiator instead of 100 mm prototype, but in final version the radiator will be 50 mm thick. Actually 50 mm radiator thickness is exactly same thickness as one of two NH-D15 rads, just that it's not one behind the other, but both conncted together, forming 1 long radiator. So the "only" difference vs. Noctua NH-D15 will be fins density and the liquid used, provided you use Noctua fans on it as well, which is what I would do 100%.

@Chimere take your time and listen to this. Very interesting. Talking about difference vs. regular air coolers technic and also about copper base plate.

Final product will have copper contact plate. Fins wiil most probably be aluminum, for cost reasons.

But don't foget that it was the thicker initial version of it + Threadripper have much much larger IHS to spread and transfer the heat than mainstream Intel or AMD CPUs. Also what differentiate them from regular air coolers is that the liquid can evaporate/dry out over time at high temps/loads.

@Chimere The company has now to find the golden middle, between radiator thickness and fins density. Thicker = lower fins density possible Thinner = higher fins density, to keep same cooling capacity as a thicker raditor with lower fins density Lower fins density = lower air pressure needed, lower fans rpms and lower noise. Higher fins density = higher air pressure needed, higher fans rpms and higher noise. Weight has also to be considered, if it's gonna be used horizontally. Great stress for mainboard PCB.

@Chimere see this