Gadgetory

despite having the worst reviews on Newegg on Amazon we suspected that the MSI 1080 TI armor card it might secretly be an actually good PCV the card seems to have a gtx 1070 cooler stuck on to the 1080i GPU and 1080i gaming ex PCB and so if that's the case we would be able to modify it and turn the card into one of the best liquid cooling candidates of the 1080 TI class as it is priced at $700 and seems to carry a custom PCB before getting to that this video is brought to you by ifixit.com and the pc essentials toolkit which can be had for $20 making it one of their cheapest yet most complete toolkits use code gamers nexus for five dollars off to bring that to 15 you can go to ifixit.com slash gamers Nexus or click the link below for more information so we bought the a 1080i armor out of pocket for 700 bucks from retailer it found that have this gtx 1070 cooler on it with maybe some very minor changes but nothing significant enough to make it suitable for a 1080 TI card and upon removing the cooler we found that it is actually a gaming X PCB a card that we reviewed and likes for the most part it's got a good cooler on it the Twin Frozr cooler on the gaming X and it's got a good PCB builds we did an analysis on our channel where he talks out all the components on the PCB they have not changed on this one aside from some minor changes to things like the LDO and some fan control stuff but overall this is a gaming X PCB for $50 cheaper which makes it sort of like a secret purchase for liquid cooling enthusiasts if you're not liquid cooling this and you're going to use it in this configuration it's a terrible thing to buy and we'll show you why through this video if you are planning to modify it however like we did with this kraken g12 bracket and just a 5 7 DLC we had lying around then it's one of the best candidates for liquid cooling that you can get because you get a good custom PCB and you get it for $700 and then you just clip this to the off to the side somewhere throw it out whatever it doesn't really matter at that point this is as close as you can get to just buying a bare PCB with no cooler attached to it obviously some disassembly required so we'll go through the numbers here we're going to talk why in its stock state this thing is so bad at what it's doing on this particular card and why all of those reviews that you've seen on Newegg and Amazon if you've looked at this card tend to be negative almost always along the lines of thermals and noise start out with stock configuration testing we have our thermal overtime charts the stock 1080p I armor cards from MSI prior to our liquid cooling mods with the NZXT g12 out of box with 1080i armor runs a GPU temperature of around 77 to 78 Celsius when left to its own devices this triggers a fan speed of about 80% corresponding with a noise level of 46 DBA we'll talk about that more later at this noise level MSI is managed to assemble one of the worst performing cards in both noise and thermal categories in our tests and they've done so only by sticking and under-qualified at heatsink on the 1080i armor if we look at vram temperatures we're rapidly nearing 90 Celsius on the measured gddr5 X module near the power component measuring the hotspot MOSFET in the center of the VRM we see a temperature output of around 70 102 Celsius not too bad let's look at a GN EQ chart to get a better idea of where things land here's a comparative chart of GPU temperatures the armor card is the worst of the AIB partner models on the bench at 77 Celsius with the Nvidia founders Edition card performing against the thermal limit point at 84 C before it starts to limit the clock this sustained at 77,000 per hours the nose output of about 46 DBA out of the msi armor card whereas the founders Edition card is running at a 45 DB a noise output with its auto configuration an 84 C temperature although this is not great it shows that the AIB partner models aren't just magically better to give an idea for efficiency versus noise here the asus strix with its 60 4.8 Celsius temperature is maintaining this with the noise output of about 30 8.6 DBA that's nearly a 2x perceived decrease in noise with a 12 Celsius reduction and GB temperature again those things don't normally coincide we're getting a decrease in noise by roughly two times as a human perceives it while also getting a temperature reduction big deal there the st2 at my CX model can be had for around seven hundred twenty dollars or twenty dollars more than this armored card and it would operate at the same temperatures as the I CX model we're seeing here with the difference being no emcee use and fixed fan speeds for both and that puts us at around sixty nine Celsius load again it's 42 DV a noise output for this temperature another non-trivial reduction over the armor the armor does okay though in some testing like power temperatures in that it lands about center of the pack as you can see in our power temperature charge this is expectedly similar to the gaming X the other msi card using the same PCB other than a couple of minor tweaks but now let's take a look at how things go with the fans fixed for a 40 decibel output on these types of devices note that we're going to be using Delta temperature measurements in this charge hence the lower temperature numbers overall that's just because they're a delta temperature over ambient configuring the coolers to not exceed 40 DBA places the msi 1080i armor card at 58.6 celsius delta T over ambient which is the hottest device by a longshot the next hottest device at this fixed DBA is the FTW 3 which runs 16 Celsius cooler than the armor for GPU temperatures at the same noise level and 11 Celsius cooler power component temperatures though they use different vrm designs and then we see another 8 Celsius reduction on the vram temperatures when measuring the same module that's placed near the power components and all of this means that even with the auto settings giving us 80% fan speeds at 45 to 46 DBA on the msi armor will still throttle the clock with the stock msi armor cooler here's a look at that throttling this frequency versus temperature and time chart is produced using the higher auto fan speed the blue line represents the stock msi armor card which gives hard right out of the gauge falling in this power virus scenario from 1630 megahertz to 1520 mega it's about a 100 megahertz reduction in the worst of times again this is a power virus so we're not a numerating the clock in the same way as a gaming application would fortunately we also have those results here's a look at superposition as benchmarked over a 30 minute burnin period as the card heats up and it starts hitting north of 70 solidus within just one minute of testing by the way we immediately drop clocks this means that we're losing framerate slowing down as time goes on and heat rises and you can see that in this chart as the frequency dips over time and temperature rises and just to really drive things home because msi often challenges criticism of their products we can look at a frame rate reduction over time in this next chart this performance was logged over about an hour of looping Metro last light benches we have more than 64,000 rows of data for this particular benchmark and that's just for the same time numbers have the thermals and we see a performance degradation as the clock drops and thermals rise it may not be a huge loss in performance nothing game ruining certainly and really nothing you would complain about in real time however it's an unnecessary loss you don't need to lose that performance seen as the gaming X doesn't experience the same performance degradation with its superior cooling there's no reason that this one should because of the same PCB just with again we're schooling so that's why you get that reduction in performance over time back to the power virus chart though we can fix that frequency fall off by putting a better cooler on there let's do it makes the most sense and use something like a crack in G 12 with an eighth tech 570 LC that we had an EVGA hybrid fan on that liquid cooler and then see how frequency responds so that's much better we're now holding nearly a perfectly flat frequency trend in fact this is almost identical to the frequency line we see with the gaming X that has the Twin Frozr cooler which we can plot now a gaming X clock aligns with the g12 NC LC clock on the armor telling us that provided equal cooling the cards are in fact equal again this is a power virus scenario but the whole point is to stress that thing and see how temperature is due and the clocks in the scenario between the gaming acts and the liquid cooled armor are basically the same bought it over time it's easy to see how much faster the liquid cooler soaks the temperature increase and how much lower the GPU fat and vram temperatures are overall the GPU temperature goes down from 77 Celsius to just over 40 Celsius what's that temperatures falling from around 71 C to around 53 C and vram temperatures also drastically reduced here's the GN EQ chart with the noise controlled to 40 DBA we're at around 20 mm on the liquid cooled version of the card down from around 50 8.6 Celsius delta T and significantly reduced on vram and pictures at once again if you're curious about noise from a stock cooler here's one last look at that this is a plot of the noise over fan speed ranging from 40 percent to 100 percent fan speeds showing a pretty stable increase in noise with each 10% rpm hike we've also got a comparative noise chart in the article if you'd like to see how this compares to other cards and keep in mind that the GPU when it is fixed with the stock cooler tends to sit around 80% fan speed and almost never goes above that instead it will reduce the frequency and as always there's a lot more in the article we've got some FPS numbers in there just to give a baseline of how the differences look on the liquid cold versus the stock cold versions not a huge difference but enough of one where again you kind of scratch your head and say why would they do this to themselves we also have some extra noise info their temperature and Pho things like that but to kind of recap things here what we have is a really peculiar video card if you're to buy this card with no intention to mod it and you want to leave that cooler on it I would tell you don't there are plenty of better options that will cool better be priced similarly or the same and you're not going to lose out on performance in fact you might have better performance because you're not going to be thermally limited and in the very least you have better noise performance which is never a bad thing now the one place that this is peculiar as I said is because unlike other products where we might say don't buy it period end of story this one still has some good purpose and that would be if you wanted to buy it and mod it so if your plan is to buy something like the gaming X because you like the PCB or you like something about it other than the cooler you could actually buy this instead get more or less the same things and the LEDs and fan controls which the fan controls are relevant anyway if your liquid cooling it and stick a looking cooler on it and you save 50 bucks off of the MSRP because you're going down we gave me next to this which is a gaming X PCB and then you can put that money towards your liquid cooling setup the one place this gets weird though is if you are if you don't have specific liquid cooling plans or anything like that then basically you might as well just buy the gaming acts between the two options of armor plus liquid cooling and gave me X because our plus liquid quotient still going to cost you more than the 750 of the game in X however we've reviewed plenty of other cards so we've got full roundups the most recent one for you to check would be our ASU Strix video card review because that one contains all the data for all the other 1080 eyes are reviewed so that would give you a full picture of everything sort of in the 750 plus dollar market but the armor card overall basically our conclusion is don't buy it if you're going to use it stock because the cooler is no good and if you want to mod it it's a fantastic candidate for the cooling if MSI could sell this thing without a cooler it seems that they would because that's basically what this already is you're there selling it with the cooler that should be removed and discarded and then you put your own on there and you've got actually a pretty good card so that's all for this one we picked up this card because of the interesting reviews on Newegg and Amazon as always if any of you see stuff like that feel free to tip us off you can tweet at us at gamers Nexus with news tips or other things review requests things like that subscribe go to patreon.com/scishow and I stopped out directly or store it out gamers Nexus net to pick up one of our shirts I'm wearing a vintage model but you can get newer ones on the store thanks for watching as always I'll see you all next time you