Gadgetory

welcome back to hardware unboxed a few days ago you might have seen our overclocking examination of the GeForce r-tx 2080 Ti well today we're going to be discussing the RT X 2080 we'll be covering everything from how to overclock these cards a look at the Nvidia scanner typical achievable clock speeds and performance if your overclocked to GPU before a lot of the things we go through in the guide section of this video will be pretty familiar to you though depending on what GPU you last overclock there might be a few new interesting things to note and of course if you're more interested in things like final overclocked performance and clock speeds you can just skip over the instructional section also if you watched our 20 atti video on overclocking expect to see most of the guide section essentially rehashed as it's all very similar to start off this video I'll be going through our standard methodology for overclocking Nvidia graphics cards the card we're using today is the founders Edition r-tx 2080 which does come factory overclocked compared to base level board partner models but the OC is quite small just 19 megahertz on the core and nothing on the memory but the card we're using doesn't really matter the steps you see here will apply to all our TX 2080 cards so the first thing we'll be doing is going through potentially the easiest method of overclocking your our TX 2080 and that's using invidious new scanner API one-click overclocking has been around for a while but this time Nvidia has built a framework for doing it rather than letting board partners create their own algorithms and according to Nvidia their framework is better trained for their GPUs more accurate and more reliable than any previous one-click methods as with the our TX 20 80 TI we're using MSI Afterburner here as it was the more reliable of the tools we used after you've installed and opened afterburner you'll see a bunch of dials and sliders but what you want to click on is the small bar graph icon to the left of the core clock slider and this stage is not necessary to touch anything else in the app simply click on the Biograph icon then click OC scanner and then in that window hit scan now you can sit back and wait a while because the one-click overclocked process takes around 15 to 20 minutes what the scanner is doing is basically running a bunch of clock speed tests at a range of voltages to find the exact voltage frequency curve for your t-80 model has its own test algorithm built-in to stress the GPU with the ability to recover from any hangs or crashes if the GPU is being pushed a bit too hard it's basically simulating what we'd do with a manual overclock but doing it faster and potentially more accurately don't worry if the application hangs or goes black temporarily during the process that's normal at the end you'll be given an average overclock but crucially you'll get a full frequency curve which potentially is a bit more efficient than a simple core frequency offset you can see at the lower end of the voltage curve we're getting slightly higher overclock so than at the top on average we achieved plus 113 megahertz but at the top-end loc is just plus 90 megahertz or so at this point would recommend heading back into the afterburner main application and cranking up the power and temperature limits to the maximum we've never been chucked a card and even a last generation Pascal card that didn't like having these limits raised all the way so if the 2080 it just makes sense to crank it up this will allow invidious GPU boost algorithm to push as high as possible on top of the frequency curve we've already set often going beyond the old limit one thing to note here is we have set the power and temperature limit after running Nvidia scanner Nvidia says the scanner only modifies the core clock so if you change the power and temperature limits beforehand the scanner might find different and potentially higher core clocks however in our experience we actually achieve lower clocks in the scanner setting the power limit beforehand so recommend cranking it up after the scanner is complete the other obvious limitation is the lack of memory clock gains Nvidia scanner only handles the GPU core and doesn't touch the memory so if you're after that simple one-click overclock solution you'll be missing out on any games you'd get from tweaking memory core overclocking is much more important for getting performance gains but a boost of memory can help out in some situations so we've got the scanner OC results now in the bag let's show you how to perform a manual overclock and after that we'll compare the manual overclock to the scanner results for this we're switching to EVGA precision x1 because it has a much nicer and more intuitive interface from manual overclock we want to start from a decent point so again we're going to crank the power limit and temperature limit to the maximum we're also going to scroll to the bottom section across to the temper tuner and adjust the curve so we're getting the maximum clock speed and all possible temperatures from here it's all about adjusting the two main sliders from memory and core frequency the basic steps are we want to increase each value by a reasonable amount apply the overclock then validate it in a program like 3dmark to ensure we're not getting crashes at those settings the amount you increase the values is up to you but note that you're not going to damage your car by choosing your value that's too high instead you'll just crash your system and have to reset it so not a really big deal so for the twenty eighty we started with around a plus one hundred megahertz offset on the core and plus 650 or the memory those are fairly conservative figures for this card that worked in 3dmark so he pushed the core up to plus 130 and memory up to plus 700 if you want to play it more safe we'd recommend only changing one of those values per test run but again the more you become familiar with the process the more you can adjust at once with plus 130 on the core our twenty eighty crashed in 3dmark so it's now a process of finding the exact limit we step down in ten megahertz increments eventually discovering that plus 120 with unstable but plus 110 was perfectly fine from here we can try to push up the memory even further using a similar sort of process but in the end RM plus 700 was right on the limit so our final stable overclocks for this card were plus one ten megahertz on the core and plus 700 megahertz on the memory that's a fairly typical figure going on what we've seen from others note that if you have a non efi on non factory OC card the core offset you'll need will be higher as you're coming from a lower styling point once you find what you think is stable overclocks again it's always a good idea to validate them further in a game running for several hours something very GPU intensive while three max times Phi gives a good indication of whether a car will work at a certain frequency sometimes it will only crash under a longer test so it's better to test both a quick note on voltage and videos training cards are voltage locked in other words Nvidia does not expose proper voltage controls to the user like you might get with overclocking a CPU on a motherboard instead we have a voltage slider but it's not really an offset well if it really does is raise the voltage limit by a few steps but still within what Nvidia deems safe in practice though raising the voltage slider did absolutely nothing for our overclock so it's not even worth using our final results from manual overclocking did give us slightly higher clocks around plus one 10 megahertz on the core compared to plus 90 megahertz we also overclocked the memory which the OTA scanner does not handle whatsoever and that provides an Adu performance boost instead of the Tomb Raider for example the results from using just the scanner gave us a four percent performance improvement over the Effie clocks Rosa Manuel overclocked Brewster performance by eight percent however one easy trip for those used in the OC scanner might be to use a fairly conservative memory overclock for simplicity's sake something like plus 600 megahertz going with that changes the results makes the scanner plus memory overclock seven percent faster than the FA clocks but again the full manual overclock provides a slightly better result so now we've shown that our manual overclock is very slightly superior to the results with the OC scanner now let's take a look at a handful of games to see how that overclock compares to both stock and felonies additional performance in shadow of the Tomb Raider running 4k with the maximum quality settings and SMA 82 X we saw overclocking provide 11% more performance than the standard clocks for the 2080 and a percent more than the founders edition clocks that's an okay result not quite as grew the overclocked results for the 2080 ti but the 2080 is clocked higher by default in both chips appear to have the same sort of clock speed cap in Assassin's Creed origins at 4k ultrahigh preset the margins are a little narrower 10% performance increased over the default clocks and 7% over the founders edition in Far Cry 5 we're looking at 11% over the default and 10% over the fairness edition in Middle Earth shadow of war there was a 12% improvement over default and 11% over founders edition and finally in Batman Arkham Knight 11% over default and 8% over founders edition so from those five games we achieved a performance improvement of 11% on average compared to stop clocks and 9% compared to the founders edition clocks we're pretty much running into a power limit here looking at the design of the power delivery on the FE cards there might be a bit more Headroom available if Nvidia properly unlocked voltage controls but this is around as good as we're getting for what we can configure at the moment in terms of temperatures and fan speeds again what Steve found in this initial review of the 2080 is spot-on compared to what I achieved on my test bench core temperatures increased from 76 degrees Celsius under load with the founders edition clocks to 81 degrees Celsius when overclocked that's a bit below the cards in native 84 sea temperature limit and well below the increased 88 CD limit so it's pretty clear that's not going to be a limiting factor for performance fan speeds as well twenty sixty rpm with the fa clocks 21 a TRP when overclocked using the default auto fan speed so a slight increase there but there's really no difference to the sound profile the card is pretty quiet went overclocked also of course both the noise output and temperatures will be different with board partner models and if you're wondering about power consumption while there wasn't a lot of difference in load power draw between the default and founders edition clock so that shouldn't be a huge surprise consuming the FE clocks only slightly faster but when overclock power draw increases by around 13% on average compared to what the FE clocks draw so a 13% increase in power consumption for 9% more performance again that's a fairly typical result if you're wondering about the actual clock speeds I achieved I was averaging around 17 70 megahertz in shadow of the Tomb Raider with the stock clocks 1865 megahertz with the FA clocks and 1990 megahertz went overclocked with the peak around 20 10 megahertz one last thing you might be interested in are the clock speeds we achieve when the fan speeds are set to maximum which could be an indication of how liquid cooling might benefit the 2080 however results are a bit disappointing dropping the card by only 20 degrees Celsius to around a tick over 60 degrees under load I only gained around 50 megahertz to sustained clock speeds which was a 2% performance increase over a standard overclock from multiple times the noise output of course overall that was the difference between 8% and 9.5% more performance over FA clocks so not a whole lot for a massive temperature drop so that just about wraps up this look into 2080 overclocking with the cards we have on hand they use some extra performance to squeeze out but again nothing too different from our 2080 ti in fact I was able to reuse a bunch of stuff from that video which is nice makes our job creating this content it'd be easier sorry if you saw some of those sections repeated if you've watched both videos and again the Nvidia scanner API seems to be quite ok especially if you combine it with their conservative memory overclock if you're interested now on going in video RTX coverage subscribe to get that content in your inbox and be sure to hit the bell icon to consider supporting us on patreon to chat directly with us about our TX cards in our discord chat or really anything else and I'll catch you in the next one you