Gadgetory

what's up guys Jays $0.02 here and that's been quite the discussion taking place on Twitter and in the comments section on my 1070 video regarding open air test benches like this guy right here oh jeez he doesn't really know how heavy this thing is versus real-world testing in a case like you see behind me on graphics cards and whether or not it makes a difference it's quite the split decision with I'd say about half the community saying the case is not the best way to do the testing because there's too many variables and then my position of allowing the open air test bench to be the best way to do the testing because then the only thing limiting the temperatures of the card is the cooler itself which is the only way to do a complete cooler test so what we're gonna do today we're gonna do a single graphics card we're use my MSI gaming x 8g a 1080 card we're gonna put it in the test bench then we're gonna put it in the s3 forty case here with two intake fans two exhaust and a front mounted radiator and see whether or not the temperatures go up in a closed case it's not to be good the new z170 classified motherboard from EVGA features eight phase PWM four-way SLI support along with top notch components to offer gamers and enthusiasts more of what they want badass motherboard with no compromises click the link below to learn more now before we get into today's test video you guys have to be aware of two different types of coolers on the market you've got the reference style cooler with boat which both AMD and NVIDIA use where it's got a turbine here that's gonna pull air in and push it out and it's not going to rely on the case whatsoever to exhaust any of the heat it just relies on the case applying air to the card so that it has air to basically suck through right here and blow it it sucks and it blows and on the flip side you've got a card like this with multiple fans pretty much all the board partners come out with the custom coolers like this which takes the air it swirls it around in the heatsink it sends it off into the atmosphere or in the case in this instance here where the case is then responsible for taking the heat and sending it out into the room so the argument has been I'm gonna leave reference cards out of this reference cards our best case scenario when it comes to case testing again as I said it doesn't rely on the case to exhaust the heat but when it comes to a card light this are you actually my computers went to sleep I swear it didn't die so the argument the community recently has been whether or not GPU testers like myself who do testing in an open air test bench are doing the community a disservice by not showing what temperatures would be like inside of a case so what we're gonna do today is we are gonna take my MSI gaming x8 G which is a 1080 card we're gonna put it in the open air test bench go through 20 minutes of looping heaven benchmark with settings max and then we are going to look at the numbers see what happened with GPU boosted at thermal throttle what happened to the temps what was the fan percentage speed and then we're gonna take it put it in the S 340 case here which has an front mounted radiator at H 100 I to intake fans and then it's got two exhaust fans and we are going to see with what I consider to be one of the worst-case scenarios here with a front mounted radiator and a closed mid tower case whether or not the temperatures go up or they stay the same or they go down in a case like this some arguments have actually been that cases Canon crease the efficiency of cooling in a graphics card and bring the temperatures down so I guess we're gonna find out whether or not that is the case I say the case but I mean the situation you you get what I mean this loop has been running now for exactly 20 minutes and I want to show you what you're seeing here it's kind of hard to see because of the you know the gray buttons behind there but we'll look from left to right the graphics card is currently running at 72 C in the open test bench with fans blowing directly on the card even I want to point that out ninety-eight percent load on the GPU that's gonna bounce between 98 99 57 percent fan speed eighteen hundred and thirty-five megahertz core clock that's kind of a sneak peak for you guys that are waiting for this msi review I promise it's coming soon and you'll see why I was taking so long when I do the review and the core voltage is running at a constant one point zero six two volts so I'm gonna go ahead and whip seroma I'm go ahead and exit the test right here and we'll take a look at the chart we're gonna do this all in one clean cut here because I don't want anyone being like out there's J being all kinds of shenanigans and stuff no no shenanigans here so here is that test run right here for 20 minutes now let's go and start but start at the the top right here the GP temperature minimum of 43 doesn't really matter cuz that card idols without the fans on but the GPU temperature here set as you can see as we go across the chart here 71 72 73 72 71 72 71 72 71 69 71 72 71 so you could see it is between 70 and 72 in fact I'm gonna say between 71 and 72 we saw 70 once and 69 once in the middle of the test fan speed highest it went to what's 57 2 % fan speed so you can see right there 56 57 it's pretty constant and then our core clock as you can see again a constant 1835 just all the way across the test it started out at 1873 and then 1860 and then 1835 and that is where it stayed and then voltage as you can see right here was pretty much a constant 160 to 1 500 and then 160 to the entire rest of the test once temperatures and everything equalized so there you go there's our baseline test here with the open air test bench we're gonna go ahead take the exact same graphics card all factory settings I'm not touching the damn thing and we are going to stick it inside of the NZXT razor case that I've got which has to 120 millimeter intake fans and to 120 millimeter exhaust fans with a front-mounted radiator on a Corsair h 100 i so i already showed you guys that's the test rig we're gonna put it in there and we're gonna see exactly what happens I don't know maybe it's gonna be worse maybe it's me better that's the whole point of this video so let's find out so the s340 is set up over there as you can see and the first thing I want to show is actually how our idle temps are lower in the case and then we're in the open air test bench because remember the msi cards and the Twin Frozr six do not turn until the graphics card gets up to 66 degrees Celsius so the fans are off right now and you can see we're idling at 39 where we were idling at 46 in the open air test bench obviously that's because the case is supplying some extra cooling to the graphics card which is actually one of the arguments some people have cases might actually help your graphics card perform better and better temperatures then open air test benches but anyway as you can see right now we've got all the settings are basically stock right now for the graphics card one thing to bear in mind is the test bench is the next 99 system and this is AZ 170 6700 K system I don't anticipate there being any difference it's the exact same driver exact same version of MSI Afterburner so I don't think anything's gonna change you can see here the factory core voltage curve everything is exactly as it ships so I don't think there's gonna be anything different but I just want to point all of this out before the test you guys can see what the settings all are that way you can see nothing's gonna change here but anyway here's what we're gonna do we're gonna load up heaven benchmark running all the same settings as the test bench it's got tessellation and extreme we have got anti-aliasing at 8 X and we are gonna run it at 1080p and we're gonna let it run for exactly 20 minutes and then we're gonna compare the results between open air and it closed mid-tower s340 that also has a front-mounted h 100 i and fans limited to 60% fan speed right there and not letting them ramp up to 100% I'm just trying to kind of control the environment here to see you know if it's really gonna have a huge impact on performance of the card enough to where people would be really upset that they put it in their case and it results were so far different from a test bench anyway when that said we're gonna start the test here and I will hey look my shadow hey I will see you guys in the in 20 minutes we're back it's been 20 minutes now I'm going to show you some numbers up here and just to show you before I do that because I know there's some tinfoil hat where is in the audience they're gonna be like you switched you're not running the s340 there it is right there the test rig is actually down there on the the floor as you can see so it is not hooked up but here are the numbers so here's what we got so far let me lock the tripod here I'm trying to this all one shot I really don't want anyone to be like you're trying to fool us so we are at 71 see right now and it spent most of its time bouncing between 70 and 71 hitting 72 a couple of times fans at 55% that's 2% slower than it was running inside the testbench 1835 megahertz 1.0 4 3 volts that's a lower voltage and it was doing on the test bench I'm not entirely sure why given this is a z170 like I said so I don't know why the voltage is a little bit lower same exact megahertz though and the temperatures as you can see is actually a couple of degrees cooler than we were seeing with the open air test bench which hit a max of 73 and was sitting around 72 constantly this is anywhere between 70 and 71 so let's do this here real quick go ahead and end the test and let's look at the chart okay so here's the chart going let's go ahead and start at the top here GPU temperature max again 73 it hit that I know a couple of times during the test but if we move our mouse cursor over here you'll see we sat most of the time at 71 drop down 67 71 69 71 72 71 68 71 70 71 71 goes all the way down to 69 right there huh 69 losers so 71 all the way toe it towards the end I can't even find where it peaked at 73 I think right there no that's 72 anyway so the temperature as you can see let's just go ahead and for the sake of argument call this dead even with the open air test bench if we take a look here at the core clock as you can see 1835 megahertz all the way across and then if we come down here to core voltage 1.0 four three all the way across the chart that did not change so I did it with lesser voltage I can't understand why it would take 1.0 five O or 1.06 two is we were seeing in the open air test bench I'm not entirely too sure what that's all about okay so there's one more test I want to do here real quick because if you recall I did this with the case fans at 60% now I've done this test already with the fans at a hundred percent it made no difference the temperatures got to the same right here so if I ramp these fans up to a hundred the temperatures stayed exactly the same they did not come down at all but what if I drop these case fans down to say 20 percent I'm curious as to what's gonna happen on the GPU temperature but I've already shown you that with the fans running at a respectable 60 percent not too noisy real-world situation here without you know being short of trying to literally choke the system as I'm doing right now with the fans at 20% I expect there to be no difference between test bench and case as you just saw alright well there's the graphics card test going right now with the case fans at 20% so you can see it's bouncing between seventy seventy one and it's been running for fifteen minutes so I I think it's time that we put this argument to bed a little bit now is obviously more that we need to talk about here and we will we'll talk about that we're gonna we're gonna close out here but this is the setup you can see it's still going atmospheric style card we're colliding on the case to exhaust all the hot air with the fans barely running listen stop making any noise listen I'm gonna unclip my mic watches in fact for those curious the fans were only running at 586 rpm and 639 rpm yeah the ships barely turned them in so there you go guys I've actually spent about the last four hours doing the testing in this video that you guys just saw I did it multiple times and then I made sure that I was having as little contamination as I could making sure the room was cooled off before starting the neck test I made sure that I had real-world scenario going here guys were mad when I said I turned off the AC and I closed the door because again that could affect the results so I've got the overhead fan going you might see the light bouncing a little bit and I had the air conditioner and stuff bent and you know open as I normally would while gaming I tried to make this situation as realistic as possible the numbers are what they are the case did not affect the temperatures whatsoever but the one number you can't argue it and I don't know why this is happening still is why the open air test bench required more voltage to get the same numbers when GPU boost was controlling things then the closed case so if anything that makes it seem like my open air test bench results being 1c hotter and requiring more voltage at least in the case of the MSI the open air test bench was actually the worst case scenario can't figure that one out you guys also saw that when I slowed down the s 340 fans all the way to 20% which the fans barely started at in fact one of the fans had to flick it with my finger to get it spinning because it wasn't enough startup voltage still saw no difference whatsoever when it came to the temperatures of the graphics card and the number to substantiate that is the fact that the fan speed when the fan percentage did not increase between those tests it stayed about 57% when it came to the max fan speed at keeping the card cool so obviously at the temps it stayed the same but the fan speed jumped up to say 60 65 % then that obviously could say well the case was introducing a situation where the fans had to go faster to stay cooler that wasn't the case I already showed you guys that where it wasn't thing see when I say it wasn't the case on a case video like this it's always so confusing that wasn't what happened here the fans ran the exact same speed gave us pretty much the exact same temperatures again the open air test bench was like 1 to 2 C hotter than the closed case so people people saying you know that the closed case would be the worst case scenario at least in my testing here wasn't the situation at all make with that what you will I think the numbers today have shown that yes I think my testing is fair I stand by it and nothing's gonna change the numbers speak for themselves and obviously as you know as a fan of reference coolers these are a lot less affected by a case than open air coolers like this guy are gonna be so I don't understand why why the big ruckus but anyway I've done my video now this is my official response to the community wanting to know open air testbench versus cases and not only that if I wanted to do case testing there's not a single case I could choose right now that would it be realistic for every single case that's out there on the market and no matter what case I chose and I'm sure someone watching this video was pissed off that I used in s340 but hey we used a case a small one and slowed the fans down to practically nothing and got no difference whatsoever on the results so that's it yes I stand by my numbers it's why I do my testing and it's the way I'm going to keep doing my testing share this video with someone you think it might help most of the time it's not gonna make any difference but I feel a little better for making this video because those that were willing to listen to the information hopefully learned something at the end of this video as always guys thanks for watching I'll see you in the next one