Gadgetory

welcome back to Pauls harbor guys this is my testing video for this tiny little two thousand dollar ish system that I built in the doctor's Abers sentry 2.0 all covered in my build video so check that out if you missed it this is part of my monthly builds series at the beginning of every month I put out some builds then I build one of them and then I test one later on in the month so I have some specific goals here with this build and that is to see how hot the CPU gets see how hot the GPU gets because I think that's going to be the major point of contention since we really have a very limited amount of space to work with here when it comes to cooling performance and when I parted out the system actually chose to go with the harder option when it comes to CPU cooling by going with an all-in-one liquid CPU cooler the Corsair h-60 here was that the right call or would it have been simpler and even a better choice to have just gone with a low-profile air cooler like the cryo rig c7 here that would give me a lot more space in the graphics card side since I wouldn't have this 120 millimeter radiator to mount so that would also mean that I could possibly upgrade my GPU from this 2070 to maybe something like a 20 80 TI but which is the better choice cooler masters SK series mechanical keyboards help you keep a low profile with ultra flat double Schottky caps and a slim minimalist design available in full size trim as the SK 650 or the even more portable 10 keyless sk 630 these keyboards feel great thanks to genuine Cherry MX low profile mechanical switches they've also got a durable brushed aluminum top plate and a detachable and braided USB type-c cable so click the sponsor link in the description for more on the cooler master SK 630 and SK 650 so for testing the system we really need to answer two questions one is how hot is it going to get and two is is it going to be able to maintain clock speeds both from the CPU and the GPU when it starts to get warm because of course heat dissipation with the case of small is going to be our biggest concern so I have three tests that I'm gonna run one is a CPU stress test with I 264 one is a GPU stress test with MSI combustor and the third is a more practical real-world test where we're gonna be gaming with Apex legends as well as streaming on the system at the same time and when I say streaming I mean we're gonna be using OBS to take what's on screen Riaan code it at a lower resolution and we're actually gonna be capturing that straight to the system rather than actually live streaming but when it comes to the CPU performance it's effectively the same thing so I first ran the with the system pretty much running its talk out-of-the-box I didn't do any tweaks to it at all so for that we were hitting mid to high 70s on the CPU during the I 264 stress test we did hit a DC on one CPU core but that's within range we were at 4.3 gigahertz on all cores all six cores on the 8700 K I did dip down to 4.2 gigahertz we were at 4.7 gigahertz max on a single core when I switched over to the msi combustor test for the GPU I was running in at 1080p in windowed mode after about 10 minutes it hit 69 degrees Celsius max on the GPU and that was just the fan running at about 2100 rpm which is about 70 percent fan speed GPU clock actually throttled down to 1440 megahertz but it is common with a stress test like combustor for the GPU frequency to dip it maxed at eighteen seventy-five megahertz and those were my numbers for the initial testing setup I want to point out that the CPU fan or the fan on the radiator here is running at a fixed 4,000 rpms for pretty much all of my tests I kept at that speed to maintain consistency across all my tests and definitely not because the asus motherboard was acting weird and wouldn't adjust the fan speed no matter how hard I tried but from there I moved to overclocking because why bother putting an all-in-one liquid cooler and getting a z3 90 motherboard and a case cube processor if you're not going to overclock it at all the answer would be case q processors do run a little bit higher frequency out of the box than non case Q processors but point is I wanted to try overclocking so I tried a moonshot first of all and just did the 5 gigahertz preset and the motherboards BIOS short answer is that didn't work we hit thermal throttling pretty quickly hit over 100 C on CPU temperature and throttled from 5 gigahertz back to 4.4 gigahertz it was around 4.8 gigahertz from time to time too but honestly not a configuration that I could run long term I tried a 4.9 gigahertz overclock where I was at 4.9 I'm one or two cores 4.8 on three five cores and 4.7 on all cores that seemed like it might work especially as I started to dial back the voltage I did a negative 0.125 volts v core offsets to help curb some of the heat output which did help some but it still ended up overheating got up to the mid-90s Celsius and eventually hit a blue screen which means you should restart and try over again so what I ended up settling on for the CPU overclock was a four point seven gigahertz overclock on two cores 4.6 gigahertz on three to five cores and 4.5 gigahertz on all course so 4.5 gigahertz on all cores is what it was running at most of the time and that's about a 300 megahertz improvement over what it was running at at stock which is better but certainly not the vast improvement that we often get to experience when we do aftermarket cooling but again very very small form-factor so that is the limitation we're doing with their run in the i-264 burn test with these settings got me no throttling after 10 minutes we hit 91 degrees Celsius max on a single core which is hots but that is with the stress test vcore was at 1.19 and we were hitting 70s to 80s see average temperatures on most course so that is within reason for a stress test in a case like this and lastly I did a bit of an overclock on the graphics card set the power limit to 114 percent which is max did plus 200 on the core clock and plus 300 on the memory and I set the fan speed to a fixed 70 percent which is basically the same speed that it was running out when I did my initial set of tests so about 2150 rpms for any msi combustor with this configuration got my GPU temperatures up to 77 degrees Celsius max but basically peaked at 76 and the GPU clock speed did improve to 1652 16 80 megahertz overall peaking at 1845 so those initial tests were meant to sort of get a baseline of what the system was capable of get a little bit more performance out of it via some overclocking and then of course I wanted to give as real-world of a test as I can figure out so for that we have this monitor over here which is 2560 by 1440 and does go up to 144 Hertz so if you were a gamer and you're investing in a system similar cost to this you might be playing at that resolution and with a high refresh rate 2560 by 1440 monitor also a good pairing for an hour TX 2070 I might add then another reason why you might have gone with something like an 8700 K over saying Intel quad core would be if you wanted to do more with it than just gaming gaming and streaming at the same time six core Intel CPUs and they quarters in particular are very good at doing that the extra cores and threads do help out and it is a practical use of the extra performance that we have supposedly gotten by going with an almond liquid cooler and that CPU choice so for this test this computer is running Apex Legends on high setting at 2560 by 1440 at the same time it is capturing that gameplay footage sampling it down to 1080 and then storing it on its local hard drive it's basically the same function it would be doing as streaming at the same time it's just storing locally rather than sending it over the Internet outside a second screen connected to the system that's being captured on a separate system so we could monitor the actual performance when it comes to GPU and CPU temperatures but more importantly how did the system actually perform so when it comes to the gameplay itself we were getting in the low range 70 to 80 frames per second average was around 100 210 when he was on the ground looking around or especially inside buildings we did peak at around 130 ish fps but we weren't really able to maintain 144 FPS solid for a monitor like this that can actually hit that refresh rate the GPU temperature got to 74 degrees Celsius peak which was pretty constant and the CPU was in the mid 80s but on average was actually in the 60 to 70 s Peaks were in the mid 80s so that was still below throttling range one thing I would say when it comes to an all-in-one liquid cooler in a system like this if you're averaging well above 60 Celsius you might consider not going with an almond liquid cooler because almond liquid coolers you want to keep the actual liquid temperature down if you're running at a much above 65 C for an extended period of time that could actually shorten the life of your pump so that is something to bear in mind if I were actually using this system on a long-term basis I would probably dial back that overclock a bit on the CPU just to make sure that my temperatures were lower than that range on average on the plus side though the CPU was able to main 4.5 gigahertz across all cores without too much issue and then our GPU is hitting 1800 megahertz pretty consistently as well so we can see that a stress test like combustor is going to lower our GPU frequency kind of automatically whereas playing an actual game-like API Apex Legends 1800 megahertz is pretty reasonable especially for a GPU with a cooler that is as small as this one so now we come back to more building which is always fun this is actually something I haven't really done when it comes to my follow up testing videos but I really wanted to see if I made the right decision going with this almond liquid cooler in here so I'm going to take this apart drop in the 20 80 TI here as well as the cryo rig so in my mind I was just gonna be swapping the CPU cooler and the graphics card but with a case like this you really have to take the whole thing apart in order to get in there to swap those pieces in unfortunately though they're all fitting especially this gigabyte graphics card and since it's a triple slot cooler with three fans all the fans are right up against the intake right here which is nice and then it's actually got a decent amount of ventilation at the top for exhaust for that card - so that actually fits in there really nicely I'm going to reconnect this and then we'll get back into testing so guys we've spent most of the day testing but we have some good results to share because this is really my question is I put together the initial build and as I compared it to some of the other builds that went up because I posted my build about the same time as quite a few other people like Kyle ufd tech Linus has done done one I've seen a few other people out there as well and the question just kept coming back to me is it worth losing the space up here to put that radiator in to get better cooling for the CPU and losing the amount of space you get for putting in a full sized GPU like this one it's very happy that the gigabyte r-tx 20 atti overclocked edition was able to fit in here this is the wind force version although the fact that the fans for the card are right up against the side of the case as well as the fan on the cravaack c7 cooler on opposite side also right up against the side of the case is both a blessing and a curse it's a blessing in that they're actually able to pull air from outside of the case inside which is good since there's no active fans in the case itself but it's a curse because it's now sitting there right next to where you're here holes are so you're gonna hear any noise that it makes and while joe was playing yeah they got pretty loud before Joe even started game testing I just did the standard I 264 burn test as I had done when the all-in-one liquid cooler was installed with Enoch - a hundred twenty millimeter fan and to my surprise no not to my surprise at all we had throttling within about a minute and 20 seconds we got over plus 90 degrees Celsius throttling and then eventually just a minute or two later we hit a blue screen actually a very strange blue screen but blue screens can often be a little weird so my next step was to reset the CPU to stock speeds in the UEFI BIOS and I did leave the memory at its XMP settings so it's still running at 3200 speed and then I still got throttling even though the CPU is just running at 4.3 gigahertz across all six cores we still had throttling within a couple of minutes 14% throttling according to I 264 and the CPU temperatures when it we're in the mid to high 90s so I only let that run for about 5 minutes before shutting it down so obviously you simply cannot run this CPU at stock speed under a significant load if you're going to be using a cooler that's this small now this is I 264 so it's a synthetic test it's a burnin test so really nothing's ever gonna hit your CPU that hard that it's going to have this effect on it so you should be able to use this configuration with the CPU at stock speeds just bear in mind if you want to do any heavy lifting with it it's going to dial back the frequency and keep itself from overheating next I ran the msi combustor test on the GPU I did not overclock the GPU because it's a 28 ETI that is manufacturer overclocked and it's a 600 watt power supply so I didn't want to push the limit there at all plus it's a 20 atti so it should be able to handle what I throw at it no problem it did hit 86 degrees Celsius during the msi combustor test it ran at 1650 megahertz initially on the GPU clock but that dialed back to about 1530 to 1550 and from there we switched to the final game test again we use the same setup as before with Joe playing Apex legends at 2560 by 1440 that was being reencounter 1080 using x264 or on the cpu on the system and then saved to one of the extra SSDs on the system at the same time I had a separate monitor set up that was being captured by my separate capture system so we could see what was going on with temperatures and frequencies so as expected because we had that additional CPU load going on the CPU got pretty hot it actually hit mid to low 90s 97 c-max was what we saw on a single-core but it did maintain 4.3 gigahertz across all cores during the test so I didn't see any actual throttling happening but I would expect that to happen if we had kept playing because we only had about five to ten minutes of gameplay if you had an extended session I can easily see this hitting a hundred degrees CPU throttling which is going to affect your gameplay as well as your Game Capture the GPU clock on the 28 ET I was at eighteen seventy-five megahertz and the temperature got up to 81 degrees Celsius so what it was able to keep itself cool enough we did have a much better framerate this time around since again it is a 2080 TI which is a 1200 ish dollar graphics card compared to the 27 T which is about 500 bucks but we were hitting a hundred and forty-four frames per second by and large and if you select scenarios like when you're first diving into the map we did see some lower frame rate so it dipped down to 110 maybe 120 or 130 from time to time but most of the time we were actually hitting 144 frames per second capped so let's draw some conclusions after all of this testing and I think the thing that stood out to me the most was actually that yes there was some decent benefits to going within 120 millimeter all-in-one liquid cooler even though it was a pain in the butt to get in there and even though it greatly limits you on the length of the graphics card you can install with a high core count CPU like a six core or even if you're gonna try to drop an 8 core in here I think that's probably the route you should go if CPU performance is your main concern if it's not and especially if you're not considering trying to game and stream on the system or something you could easily go with a lower powered CPU a quad core and they could get by player games you wouldn't have to worry about CPU throttling and you would be able to fit a much larger graphics card in there also when it comes to noise though even though again we're just using the single slim 104 220 millimeter fan to cool the allman liquid cooler it was much quieter relatively speaking then the crack c7 and that's simply because the cravaack c7 had to keep its fan spinning at a very high rpm and again it's just positioned right there next to side of the case so you're gonna hear that noise when it's generated what I think would be cool to see although it would require a mod on a case that already cost 250 dollars would be for someone to mod the opposite side here to sort of cut out a gap so you could have air passing all the way through 420 millimeter all-in-one liquid cooler in that case you can probably also mod it so you could fit a full size 25 millimeter on there as well that would probably give you much better cooling and give you a better solution if you're trying to find that balance between like a really high performing CPU and a pretty good high performing GPU and still have it all fit in this very small form-factor ultimately though I think this is a niche case I think it's gonna stay that way it's only going to fit the needs of a few specific people but I think for a limited run case that's actually probably okay in the end though I think we're all just sort of limited by the laws of physics and if you're looking for the perfect trifecta of size performance and temperature or noise generation you're gonna have to give up on one of those corners with a case like this the size is definitely there but you're either gonna have to have a low performing system that you can keep cool or a high performing system that makes a lot of noise like this system did when it was fully kitted out and overclocked but thank you guys so much for watching this video I hope you've learned a little bit more about the doctors aber sentry 2.0 again I'll post links to all the parts that I use down in this video's description hit the thumbs up button if you enjoyed it and we'll see you guys next time good initiative bad judgment trying