RTX 3080 Temperature Testing... Not the result we expected

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
so i've been seeing a lot of people talk about 3080 and i remember there was a lot of rumors of people saying that fan on the back it's going to make my cpu run hot what a stupid design so i figured today we'll go ahead and test that yes i am well aware that steve already made his video with his thermal imaging showing the airflow pattern and how even he was surprised it wasn't very conical like very much of a of a sharp tunnel but unfortunately he didn't end that video with a practical test by actually testing the temperatures in a system [Music] but i can tell you i've already tested this i put this inside my threadripper system this is my system here at the studio that i use that mostly i play world of warships on while phil is working and i tested for two hours straight what temperatures were like with this so what i'm gonna do right now is i'm gonna do an a b comparison just giving you another point of source of info to use with our hot cpu that's on air cooled or air cooling which is kind of funny because that's the only air cooled cpu that i have and we're gonna see whether or not with two cards with the same identical tdp with two different cooler types actually affect cpu temperature once and for all so we need a baseline test now so here's the thing although the founders edition is a custom pcb in the evga 3080 xc3 xpx1 lmnop qrs t u v w xy whatever i probably tried to alphabet wrong is also a custom pcb not reference as far as i know but it uses a traditional style blower and that means that air is going to come out the back presumably some of those tiny little holes fins this way towards the glass that way towards the motherboard some may be going that way towards the outside of the case kind of coming out in all four directions so that'll give us a very good a b comparison because it's still a 3080 that uses 350 watts with the same boost tables and all that sort of stuff so had i done this with a 2080 ti you could be like well the temp differences are because it's not as hard of a card 250 to 290 or 300 watts depending on the custom variants at least this way they are as apples to apples as they can be so let's go ahead and get this guy in there so the system that i'm using right here is my threadripper 3960x because you have the 70. so that means this is the 24 core i believe 24 core 48 thread variant bottom line is it's not overclocked or anything like that it's just out of the box because i'm on the air cooler but it does have the be quiet dark rock pro for for the tr4 obviously and why does evga always have to have their plugs inverted backwards from everybody else but with the standard card in here you can see air is going to come out this way and probably bounce against the glass and a lot of them make it this way but it's going to the heat is going to generate heat inside of the case so the thing that people are concerned about is with the rear fan pointing up and blowing through the card that this fan is going to suck it in there's a fan right here in the middle too and it's just going to raise the temps in your system unnecessarily we lock the cpu cooler to 75 percent we lock the case fans to 60 and we lock the gpu fans to 65 we need to control the environment without letting things ramp up otherwise we can't measure what the differences are our air conditioner here in the studio is keeping things at 71 degrees fahrenheit that's whatever this is in celsius and let me flick that away so we're using superposition benchmark for this test as i already said 1080p custom yeah i know people hear 1080p that's not hard but 1080p extreme is extremely difficult to run it's way harder than 4k optimized on this and so we're going to be running it with ada64 going which is doing all of our cpu monitoring and then msi afterburner is going to be displaying it in its osd or on-screen display for us and then i'm using game mode and not benchmark because what game mode will do is it will allow me to find a spot in the room where it's under extreme load or at least where i can make sure it's at the most usage and i think 98 percent as high as we're gonna see it go and then just leave it static because although people don't sit there in their games and just stare at the corner like this but you can see now our fps is constant our gpu percentage is at 97. it bounces between 96 and 97 our gpu temp is climbing this is our cpu temp right here and what we're going to do now is we're just going to let this go for i think 15 minutes seems like a good round number all right so we're about 15 minutes in actually it would have just ended um so here's where we equalize our gpu one as high as 82c remember we're not letting the fans ramp up we've kept them static at 65 and have a feeling with how cool and quiet these fans are i have a feeling that it would have ramped up higher than that but because everything is static that chassis fans can't go up the cpu fan can't go up we pretty much equalize at about 52c on the cpu now somebody might look at that and go wow you started at 40 degrees and now you're up to 52. clearly the graphics card's causing heat well yeah we know that the graphics card's the heat source in here and it's not even the one that's in question about blowing the hot air up so this is our baseline we now want to know if with this graphics card if our cpu is going to get any hotter than that but first there's another test i want to run real quick i want to go ahead and un-choke this graphics card and case by taking off the front panel wow i felt heat come this way considering the air is blowing that way um we all know the h710i as steve called it as a brute force method of trying to keep things cool but i bet you right now if i let this run for 10 minutes with the front panel open both the cpu and the gpu will come down as a whole in fact the gpus it's already starting to come down it's now at 81 instead of 82. so i'm gonna let this run for 10 minutes and see what our our temp drops down to after the 15 minute run and then 10 minutes of an open front panel there's no filter here it's just open i can stick my finger in the fan if i wanted to like that so it's it's wide open all right so it's been 10 minutes now with the front panel off and you can see whoa i moved it no okay you can see that our cpu temp is at 49c 46c it was actually at 44. there it is um 43. and our gpu went from 82 down to 78. now some might look at that and go well clearly the gpu temp coming down lower the cpu temp what if i told you the cpu tip came down before the gpu temp the cpu temp came down almost instantly after opening the front of this case the gpu temp took quite a while to come down fun fact the gpu under load is a lot harder to pull the temperature down than it is to have it not get to that higher number to begin with jay what do you mean trying to pull it down from 82 to 78 is a lot harder than letting the fans start or let the gpu start with a higher percentage fan and run at a cooler temperature so if i were to stop the test right now with the front open like this let everything cool off and do it again it probably won't get to 78 it's much harder to get rid of that energy than letting the energy form to begin with we're going to shut this down we're going to throw in our 3080 founders edition i have a prediction that nothing is going to change at all for the cpu i feel the cpu temperatures are going to be exactly what we just saw i think people are severely blowing out of proportion that fan and what's gonna happen if anything i feel like this might help our situation because we're gonna be blowing air now directly on our ram which is not normally getting any airflow right now because as you can see our fan is in front of the ram and then this middle fan is blowing air out so we actually have some cooling oh wow that rear that rear ram right there holy cow okay that wasn't unexpected phil feel this i touched the very back one that's cool it's hot it is hot it's like why is the why is the background so hot probably because it's close to the vrm right there in that little heat sink now you can see why i keep telling people having air flow over something that's hot that has no airflow is better than nothing even if the air is warm which is exactly what i think we're going to observe here jay the cooler is what's heating it up no that's far hotter than the air that's coming through this cooler let's just throw a number out there you can't take 35c air blow it on something have the thing it's blowing on go hotter than the temp of the thing it's blowing at it you can't have 35 air heat something up to 50. it doesn't work that way you say stuff why didn't you put a space between these slots it's so stupid wow just jam my fingernail so while we're waiting for the system to boot up here let's go and take a look at the layout of the 3080 so given the size of our be quiet cooler here you can see the fan is actually farther to the front than where the fan on the gpu is so only part of the gpu fan has a chance of making it into the intake of the cpu right there because remember the air is going this way i can tell you right now though given the position of the ram right there that that air is blowing directly on that ram which now makes me wish the front fan was a pass-through too given what i just felt with that last ram stick i think the real disaster here is these cables honestly but yeah so anyway same thing we're going to run 15 minutes we'll show you the temps take out the front panel 10 minutes without it see what they come down to and then that will uh obviously paint us a pretty good picture here so we just passed 15 minutes our gpu is 15c colder at the same load i believe we're holding a little bit higher clocks and our fps is also the same so look at this though we're 3c colder on the cpu with the exact same tests in the exact same room 71 degrees fahrenheit still all of the fan speeds are still exactly the same front cover yeah front cover is on so we started talking about this for a second and i was like wow i expected to show people there wasn't going to be any change i didn't expect to come back and say hey guys everything's running cooler the gpu it's clear that the fe cooler with the massive heat sinks on it the the dual fan the single blower fan and the pass-through fan are doing something i have two theories as to why the cpu is colder i feel like some of the air blowing through onto the heatsink like i said is assisting in cooling because we know that we're causing a situation here on purpose where not as much air flow can get through the front of the case as we would like but i feel like the air that is coming through the cooler hitting the the shroud and the fins is offering some level of cooling and then it dawned on me hey jay you're stupid i had nothing to do with this it just dawned on me that i'm stupid but i realized at that point that blower fan is taking a lot of the heat out of the case whereas the triple fans like this or even the double fans all of this hot air is making itself into the case so the top fans and the rear fan have a much harder job at exhausting that hot air and where does that hot air that's below the cpu have to go up through the cpu so that means the cpu cooler as it as a whole is going to be grabbing a lot of that hot air and sucking it through as this as the gpu is heating up the entire chassis so that blower fan is actually doing a uh a significant amount of work at keeping the overall chassis temperature lower so i i mean the whole time it's in here explaining it to you nothing's changed it's it's locked out it's not getting any worse so now we're gonna see what happens it's so violent now we're gonna see what happens if we take the front off and i bet you we could see this in real time let's just point the camera here i bet you in real time we'll see just how fast these temperatures can start improving in terms of bringing down the temps and i told you in the last clip the cpu temp came down before the gpu and what that tells me at least in our last clip is the fact that the cpu was more affected by your case airflow than the gpu obviously otherwise if they came down together one could argue that the colder gpu which again the gpu didn't even come and there's that spike i'm talking about something happening in the background with the os that's going to come right back down something's loading pinging a server updating i'm not sure give it a sec there it goes see it came back down who knows what it's doing when those updates are paused but there we go so we already lost the fact that even hit 47 for a second but look we've lost 2c back up it's funny as things come down it kind of fluctuates as it bounces between those different uh sensor points see now we're solid in the 65 now we're going from 48 down to 47 you'll see that trigger itself again here in a sec but that's just in the minute or so i've been talking to you since i took that front panel off you can see the temps already starting to come down all right 10 minutes left siri set a timer for 10 minutes we'll see what it equalizes at but i think it's pretty obvious now that uh the it's been debunked it is completely bogus that the blower style cooler or whatever you want to call this this fe cooler it's going to have any sort of negative impact impact on your system in fact compared to a comparable card that was also set at a 350 watt tdp limit for this test to make sure one card wasn't boosting higher than the other you can see the fe actually kept higher boost clocks it kept better temperatures and it did not heat up our system as a whole because this type of cooler which is very traditional now these days does increase the internal temperatures of your system it's also why the blower style cooler was something i always recommended in small form factor cases where you didn't have adequate air flow because the only thing that was responsible for keeping this card cool was the air it was exhausting or through its its cooler pulley in here and exhausting there it didn't exhaust that hot air into the chassis some of it would leak and bleed out some of the edges and there was a little bit of radiant back heat coming off the back plate but as a whole this is a much better solution for something like the loki ghost s1 case maybe an encased n1 stuff like that where you don't have a whole lot of chassis fans although these sound like hair dryers they are the best thermal solution for a case like that it's if you put a triple fan like this into a small like ghost s1 type case you'll find this probably will thermal throttle simply because the case can't handle exhausting the heat whereas this actually does a better job of it and we're only a couple minutes now into that 10 minutes and you can see we've already come down to 64c and 45c on the cpu what's funny about that is the fact that because our cpu is not under any sort of significant load it's probably only at like 10 15 load while running this test at all our gpu being maxed out these hot temperatures are not affecting our cpu so i'm going to go ahead and call it here there is no point to let this test go any longer because the point has been proven but jay what about the 30 90 that's clearly a hotter card no actually the 3090 runs cooler than the 3080 in all of our tests because the cooler is so overbuilt it's 30 watt tdp higher than the fe 3080 however the cooler is definitely built greater than that 30 watt difference it can handle it enough to where the actual load temperatures of the 3090 are lower than the 3080 across the board and as such you're going to see the same thing we just explained right here happening even more efficiently because you have a bigger 100 what was it 110 millimeter fan i think yeah it's 110 millimeter fans on there so you have a bigger fan blowing air up into the cpu like i just said and even bigger fan in the front blowing air out of the case never making it into the internal volume of the of the air so with that said share this video with anyone that you see commenting that kind of cooler is going to just heat up your cpu because it's bogus it's ignorant it's incorrect so if you're new around here and you've followed our channel or found us because of all the 30 series hype once you go and hit that subscribe button it doesn't hurt for long you'll get over it quickly i promise but we're sure to bring you more content like this where we do crazy experiments and stuff and with the rip gn series being revived which is basically where steve and i go back and forth over taking control of the leaderboard and who can have a faster setup it usually gets pretty out of control and it's worth watching and steve and i have already agreed rip gn and rip j it's on but now with this card it's way too slow
Info
Channel: JayzTwoCents
Views: 1,188,237
Rating: 4.9304261 out of 5
Keywords: rtx 3080, 3080, nvidia 3080, 3080 crashing, rtx 3080 crashing, rtx 3080 issues, rtx 3080 problems, rtx 3080 black screen, rtx 3080 overclocking, 3080 overclocks, 3080 crashes, worst 3080s, worst rtx 3080 crashes, 3080 cooling, 3080 cooler, is the 3080 cooler bad, is the 3080 fe bad, is the 3080 cooler bad for cpu, cpu temp, 3080 cpu temp
Id: mJ4ha0zsWHw
Channel Id: undefined
Length: 16min 18sec (978 seconds)
Published: Sun Sep 27 2020
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.