Gadgetory

everyone welcome back to another episode of ask GN if you have questions for the next episode post them in the comments section below or if you're part of our patreon discord community you can post them in the ask Jeon channel where I'll check you for the next week there have been a lot of questions since the last episode we're gonna answer a couple of them and then I've got some sort of a monologue on the H 700 I versus the H 500 P and how product reviews work before getting to that this coverage is brought to you by EVGA and NVIDIA with the destiny to 1080 TI bundle a 1080i sc2 comes with a synchronous fan control for its dual fans and nine thermal sensors and again includes destiny to learn more at the link in the description below so let's open with sort of a monologue on the H 700 I versus aged 500 p KS reviews there's a comment I got that's there were maybe two or three of them that uh stood out and we get maybe one every 100 cows or so but this this one specifically said Karin's access I love how you bashed the H 500 p4 thermals but H 700 I was actually worse than thermals so a couple things here one that's wrong uh it's not worse than thermals as the charts show so to clear that up if anyone thought the H some 100 I was worse it's probably because one chart had a variant of the h5 was one of the best if not the best on the CPU torture thermals chart so let's clear that up first in stock to stock testing the H 500 P is in fact objectively worse than thermals than the H 700 I the gap is not that large but the difference in criticism between the two cases was that one focused on the cooling aspect for the Aged 500 PE and the other focused on the smart aspect for the H 700 I so let's talk about how we do reviews and why for for starters and start with marketing and then we'll get into the rest the marketing for the coolermaster H 500 PK s very directly stated in the review guide material sent to me that not only did the H 500 P position itself to have high-volume airflow it literally says that it guarantees high volume air flow the same guide also suggests several times over by no accident that the aged 500 P is a high airflow case and they suggest different ways of testing that on the consumer side everyone knows the H 500 P is a half successor so the case was marketed and we had review guides sent to us saying it guarantees high airflow our job is to sit as a last line a last filter between marketing and the consumer so if the marketing is this thin is high airflow my job is to validate that claim which we did and I'll talk about some of the differences in thermals in a moment NZXT is marketing was that their case has intelligence and machine learning and that it creates using adaptive noise reduction a better a better noise rpm thermal response curve then you would otherwise have so then my job is to test this thing and see if it does so you can understand then why for the Aged 700 i review we would focus on that aspect and for the Aged 500 p review we'd focus on that one the airflow I think I'm pretty sure about 95% of the the viewer base knows this and it's just still worth talking about how we review products because everyone's gonna do it a little differently I'm not here to really debate about it because it's just a core philosophy and it's not changing but basically the way I look at the products is how was it marketed how was it marketed to me before launch how is it marketed to consumers whether through direct marketing materials or indirectly through media coverage because maybe at a press event they told our team that the case does XYZ and then we of course report that in the news right up and so now indirectly or otherwise the case is understood as doing X Y & Z and so at this point we have to validate the claim when the product comes out and make sure it does do in fact XY and Z and does it as well as they've said it does so that's how I view it as for the thermal discussion with case reviews we again look at them how they're marketed if be quiet came to us with a case and they said this case is guaranteed to be low noise output you are guaranteed to hear very little of the case volume if we then test that and find that the case is middle of the stack averaged 37 decibels for our test system we're going to spend 10 minutes eviscerating it and probably start in the coverage by peeling the foam out of the case much like we started this coverage with pulling the smart device out of the case or like we started some of the age 500p coverage with popping the glass out of the panel if the idea is with each one we're looking at what it says it does and then validating it and sometimes the companies marketed their products appropriately and that's accurate other times it's not so it's not just the thermal thing trying to poke holes in the review by saying you cared about thermals for this case but not this one is sort of misguided because it misses the fact that marketing matters and like for be quiet again for example if be quiet had average thermals but it were the quietest case on the bench I wouldn't really make a big deal out of it because the point is that it's quiet and they're marketing it as a quiet case in this this hypothetical scenario so that's something to remember and also just to reiterate here the other big point with the H 700 ion the H 500 P was if you look at the unconstrained thermal performance the H 500 P when you remove the front panel completely would drop 10 to 12 degrees Celsius on the CPU thermals H 700 I would drop about 6.3 exactly 6.3 action and about five for the GPU so the H 700 i with its front panel design is pretty close to its maximum potential for the stock configuration it has with our test system whereas the H 500 P the coolermaster case was 12 degrees away from its maximum potential and then when we replaced the front panel with mesh we got it to about 8 degrees improvement so the difference there being that now you are maybe I can't remember if two to four degrees away from max panel list potential versus the mesh because you're still fighting a mesh so you've got static pressure to consider for those big fans as opposed to 12 degrees away so what we're looking at there isn't necessarily outright thermal performance of which one's better and the H 700 I was better anyway but we're not really looking at that what we're looking at is how bad is the panel design or how good is the panel design and in case a vers B if one is pretty close to max potential that's not a bad panel design in terms of airflow so this goes back to a discussion of thermals aren't some magical number that you just apply to things or gets assigned to them like a score from fire strike it depends very much on every single component in the bench how its tested what software is used to test it is a B X or not is it a fixed power load or is it modulating are using liquid or air which has an impact on how long it takes to ramp thermals what materials are the case is the case made out of because that'll impact kind of how long it hangs on to its temperature as well so yeah lots of stuff think about here in terms of how reviews work for us but I wanted to point those out because it seemed like people were upset that I hurt their feelings about the H 500 P in a couple cases it was not that common but it was common enough that I thought you know we should talk about how we review products so people were upset that the H 500 P didn't get a good review and you know the H 700 I didn't necessarily get a good review either we just said it's a good case that they ruined with this and the age 500p really had a lot more problems than the airflow so that's the other thing I mean like it panel quality is very cheap it doesn't innovate features it doesn't do anything impressive and new and it countered its own marketing so that's kind of the difference it's more than just airflow as the point here it's everything and age 500p advertised airflow so it got a lot of focus but it also had a lot of other flaws it still looks okay but like we said lates 500p review you know if you like how it looks great go buy it but uh otherwise there's all the information so just want it to address how a review works from my perspective so let's get on some questions the first one is from discord Unilever who said Steve you say about this you say about some games that in overclocked GPU doesn't work well with it what does this mean exactly what happens if you do have an overclocked GPU and what would cause these problems I said this in a 10-7 TTI review to give everyone some a point of reference so for honors specifically that game there are a couple others for honors specifically behaves very oddly with clocks it's not a clock fluctuation or modulation what it is is we can apply a 225 megahertz offset to the colorful 1073 I reviewed and have it stable in every game that we test otherwise fire strike everything like that you play for honor and what happens and to some extent Ghost Recon both Ubisoft before honor with a two 25 megahertz offset as we've shown in the past with this particular card is micro stuttering it has so bad that your frame times just have huge latency between them so you end up with something like an 11 0.1% low value now a converted chart and that's happening I think because for honor seems to enumerate the clock a little bit higher than other games so I think having those bigger OCS on it pushes it beyond the cards capabilities in a way that other games don't - they don't they don't seem to create that scenario so it's a very specific game for it so the 1070 SC I've thought my head we had a much bigger overclock for other games than we dropped it to something like a 25 50 maybe 50 or somewhere in that but a sub 100 megahertz offset and that was because it had way worse performance otherwise and for honor now if you look at our Ghost Recon wildlands charts from the 1070 TI review you'll also notice that the 1070 TI when overclocked although it has higher averages actually had lower 0.1% low values the overclock was not unstable I think that's a misconception that's pretty common it was not an unstable overclock and in fact the clock was sustained almost perfectly flat so that wasn't the problem either but because of the game works we're running into some other kind of limitation I don't know what that one specifically is but sometimes it can be power or your power throttling really hard on occasion although you tend to see that reflected in the clock rate or it could just be that you're on the precipice of instability with that game to the point where your frame times look worse but your average looks better which goes back to why we do things with frame times so yeah basically you just keep an eye out for micro stutter and for what what people would call stutters in general and your average might be higher but behind the scenes it might be actually worse performance overall and it's just on a game by game basis and I don't know why it always happens but next question is from Shiva who I think posted on YouTube and said hey Steve question for next asked yen since placing a iOS at the top of a case as exhaust is now common practice would it be practical to use the rear fan as intake in theory it should help with vrm cooling and keeping a positive air pressure and you could probably use a magnetic air filter thanks so yeah I I'm gonna take this question modify it or modify the answer to be a little less direct about that we've done testing with top intake and in the past years ago at this point for the most part and you can actually end up with better performance I know you said rear end tape but with tobot and take you've got something directly above the cooler if you have a tower cooler and above the vrm you get air over both those things your problem becomes dust of course through the top but hopefully you can deal with it with a filter rear intake I can't say that I've actually done any objective numbers testing on that the problem is I guess well if you have a CLC you avoid the problem of creating a vortex or conflicting airflow so you avoid that problem but yeah you would get air over the VRMs it's gonna be cooler air than what the VRMs are generating for a heat output but um or power output I should say but I I think I would try and do a top intake instead if possible unless it's that slots taken by your CLC because if you kind of think about it well the times the backfoot case for people is really close to a wall it's not gonna have the most airflow there the most access to air and a GPU being lower in the case is if it's a blower card especially but generally there's heat coming out from the back of the case that's gonna be lower than the top rear exhaust would be so you've got power supply exhaust you have GPU exhaust and those things will kind of just get sucked back into the system and that air is still gonna be cooler than what's in the case most likely it's diffusing as soon as it exits the case but I think optimally you do top and take if you're concerned about be around temperatures well have more on vrm temperatures very soon next question actually I should I should mention this too before people fight in the comments over it yeah heat rises yes but not really a relevant amount we've talked about this before so there are reasons to do top intake or exhaust that do not have to do with the fact that heat rises because heat may rise but as we said in an old old SGN it's relevant when you're blasting air with a thousand p.m. fan it's just it's such a small amount of of heat rising from natural physics that I would not let that factor into where you're placing your fans you put your fans where you need airflow and that's the end of the story don't worry about how heat moves because it's it's really not going to play into it next question Gorn why don't we see broader use of USB C in the desktop enthusiast market seems the trend in the laptop world both PC and Mac so is there some reason why USB sees promise of one connection to rule them all shouldn't take off for desktops I think I don't have a great answer for this but I I think that we're seeing it more in motherboards now for Kate you really do see it in cases though I think for cases to start implementing USB C required a few things one is they needed motherboards to start implementing it it's a chain reaction they can't really lead that charge they need the motherboard vendors to lead it and to the other thing to consider is that cases have often a multi-year production cycle and a lot of companies kind of by the older used tooling from from companies that are bigger than them from OMS and suppliers and those things aren't necessarily updated for several years at at the RL 0 5 to Lynne is still used like 4 years later so it'll take a while for the cases to kind of catch up because things that are on like a multi-year production cycle it doesn't seem like it's hard to change an ad USB type-c in there but if they have to change like the molding or anything like that the cost might just be prohibitive or maybe they've already started production I think you'll see more of it towards Computex this year that should kind of be around when the cycle refreshes again next question Dave J regarding liquid metal Tim have you looked into corrosion concerns I've read that the gallium alloy used in liquid metal reacts negatively with most metals especially aluminum and copper potentially causing damage and poor performance long term have you considered testing this out yourself and is looking safe over long term for dewetting cpus i have not tested it myself I can tell you that some liquid metals like conductor not should absolutely not be used with an aluminum heatsink because they will destroy it they don't corrode so just read the read the package they should say when there's a metal that they definitely don't work with conductor not makes it really clear in their their materials that come with the liquid metal that you absolutely should not use it with aluminum I haven't tried the other liquid metals the cool abs I haven't really used so I don't know how those are for long-term use but I yeah I mean the only thing I can really speak to here with any level of authority is that just don't use them with aluminum coolers copper I'd I don't directly I have not used any that corrode copper but we'll see if that comes up the future next question is a rythmics just out of curiosity it would it be possible to test CP to GPU or vice versa contribute a contribution to contaminating each other's thermals as most of us already know poor air flow increases internal temperatures which effect in effect increases CBN GB temperatures as a result my question is how much do they impact each other long story short yes so our case reviews the torture tests our CPU and GPU burnin and 3d mark is a lighter version of CPU and GPU burnin and then you can look at it in a blender tests instituted this year where we do a blender CPU workload that's only on the CPU and a blender GPU or low that's only on the GPU and that will help you understand how the to impact each other but yeah if you're running a torture on each it gets a lot hotter hence the name torture test because obviously if a GPU is spewing heat inside of a case the CPU is gonna keep pulling that back through its own cooler so we do actually account for that in our case reviews just if you didn't already know that that's why all those different charts are there because they tell you slightly different information a couple more maybe one or two more here let's see let's do this one's fast commander Powell what's the deal with faulty tr4 sockets and how can the consumer avoid them we have this problem at PAX when we did a coolermaster c7 RP build basically take your Torx wrench that is provided with the thread ripper CPU locate the Torx screws that are on the flank outer flank of the socket not the one for the CPU mount but the outer ones and then tighten them until the wrench clicks they are the same torque spec as the CPU that's how you know if yours are loose or not and that's how you avoid them you just before you build the system tighten those down I actually have had to do it to a couple boards myself so next question is uh let's see this one is very fast as well cuz I'm going to be reading an answer from build Zoid Samuel CC said longtime listener first-time backer is on patreon thank you the question basically is what's the difference between fixed voltage and adaptive voltage for overclocking and that that's kind of the crux of it so I sent I sent the whole question over to build Zoid bill joy said both me and Roman that's their Bauer have done tests on idle power draw using fixed versus adaptive voltage you save like 10 to 30 watts depending on the CPU and then he says actually last 10 to 30 watts at the wall and he also says basically it makes little to no difference on mainstream CPUs on a thing like the Intel high-end desktop platforms it makes more of a difference so that's what I've got for you on that one I use manual voltage personally so I actually haven't really messed with adaptive last question at all all's ati eighty I lols says important question do you have more cats yes I don't know if we took a shot of midnight not too long ago if we can find it we'll put it in thank you for for watching as always you go to patreon.com/scishow and I stop site directly or join the patreon backer discord stored on cameras Nexus Don has a pic of a shirt like this one or one of our new stickers and subscribe for more I'll see you all next time