Gadgetory

today we're reviewing the r-tx 2060 with additional tests on if an RTX 2060 has enough performance to really run games with ray tracing basically battlefield at this point and that's on the TU 106 GPU we have a separate teardown going live showing the even more insane cooler assembly of the RT x 2060 over the 20 80 series besting the previous complexity of the already insane twenty eighty series today's focus however will be on performance in gaming thermals RT x performance power consumption and acoustics for the Founders Edition and Vidya RT X 2060 before that this video is brought to you by Thermaltake score p3 case the core p3 is one of the most unique cases on the market it can serve as an open-air standing chassis a test bench in vertical or horizontal orientation or as a wall mounted showcase PC the core p3 now comes with a 5 millimeter thick tempered glass panel for its side but keeps the front top and back open for air the core p3 versatility as a display piece test bench or standard desktop is reinforced by its price of roughly 110 dollars on Amazon you can learn more at the link in the description below we're gonna get right into this one today because CES is now when the video goes live and so everything's a mess including having still our new copper Mule mugs with thermal conductivity for copper on them on the table so that the 2060 is $350 that's first they need to know there's no more special founders Edition price it's just 350 flat and that is for the the the founders Edition and for partner model baseline MSRP partners can go over that you'll see a lot of cards for 400 but 350 prices will actually exist for some of the cards just like the 27 tees have there's low on models like the EVGA black for 500 so that's what you need to know there's still the a designation for the special GPU thins the higher frequency ones which we showed previously that still exists with the 2060 as far as we know the 2060 f.e is an ace Q it has been a bit higher than some of the ultra cheap cards which we might see later so other prices for cards just to give our own perspective the 1060 these days about ten dollars us for a 6-ski by model if you can find it some of the cards are more like 240-250 depend on where you look which was baseline MSRP but we do see them more regularly than not for 210 or so these days the artists 590 is about 250 to 270 some at 280 but on average you'll see them 260 or to 74 590 which does generally outperform a 1060 but not always GTA is an example of what it doesn't but most the time it does the r-tx 20 7500 bucks baseline 550 on average and then the 2060 falls in between now for performance we're gonna do the content in this order so there's games for DirectX 12 and DirectX 11 first then we'll talk RT X in battlefield 5 we'll talk thermals extensively we'll talk nose and we'll talk power consumption and overclocking and then we'll give you the wrap-up at the end Sniper Elite 4 is our first text this one was chosen because of its well implemented low-level API access and bypass to the normal abstraction layers presented by both rappers and dx11 we'll start with our 4k test for a baseline with the most data to compare against at 4k Sniper Elite 4 has the stock r-tx 2016 56 FPS average with lows at well timed at 50 fps and 49 fps 1% and 0.1% low respectively this is indicative of consistent frame time performance something we'll look at next overclocking the RIT X 2060 to a 160 megahertz offset because 175 was unstable in this game landed it at 61 FPS average a gain of approximately eight point nine percent not bad the EVGA r-tx 2070 low-end model for perspective performs at 64 FPS average which is only 4% ahead of the overclocked r-tx 2060 granted overclocked in the 2070 furthers its lead to 70 FPS average but the point is that the 2060 is nearly achieving 2070 levels of performance when the former is overclocked the RX 590 ends up at about 43 FPS average when stock allowing the 2060 elite of about 31 percent over the 590 it also has a price difference but we'll talk about that in the conclusion the positioning of the 2060 puts Vega 64 just barely ahead of it and functionally tied with the 2060 overclocked the 1070 TI also equates 2060 levels of performance with Vega 56 distant when stock generationally the r-tx 2060 outperforms the gtx 1066 gigabyte gaming x and is 37 FPS average by 52% that's a massive lead and some of that is because of the twenty-sixth ease improved capabilities with asynchronous command queuing type really 4 also benefits from the boosted memory bandwidth on the 2060 although the 50% increase in lanes on the vectors further helps over the gtx 960 SSC 4k perforins has improved measurably from 20 FPS average 256 FPS average higher resolution throughput has seen heavy focus over the past two generations well look at frame time performance next as a reminder frame time charts are the most representative look at real experience displaying a frame to frame interval for every instance of gameplay this helps us look at data that would get averaged out in bar charts and best illustrates performance for each individual frame present lower is better but more consistent is more important than lower starting with a plot of the gtx 1066 gigabyte gaming x from last generation we see a line plot that sits between 25 and 32 milliseconds with overall frames of frame pacing deviating no greater than 1 to 2 milliseconds per frame the frame rate isn't that good as we saw in the previous chart but the frame time consistency is about as good as it gets lowering resolution resolves the frame rate issue but frame time consistency is the more difficult aspect to get correct and this one already got it correct the RX 590 plots next show an improved frame rate by way of a 20 to 27 millisecond frame time average also nearing peak consistency the gtx 1070 plots better frame rate than both of these nearing 20 milliseconds average and is the best performer up until we plot the our TX 2060 the 2060 s sits between 15 and 20 milliseconds on average with sixteen point six 67 milliseconds representing a 60 FPS throughput if you had sixteen point six six seven across the entire distribution the 2060 like the previous cards has overall excellent frame time consistency and would present minimal hitching or stuttering in this title 1080p should be more interesting from a generational scaling standpoint the our TX 2060 ends up at 151 FPS average stock which scales 46 percent over the 1066 gigabytes stock card or 165 percent over the gtx 960 SSC the 1060 scaled 82% over the gtx 960 SSC for perspective and the difference is that the price has moved with the RT x 2060 whereas the pricing of the 960 and then 1060 we're pretty similar at launch the 960 s launch price was about 200 to 240 dollars to banana partner models with a 10 60 at about 240 to 250 and the 20 60 at $350 F 1 2018 give us a look at a well built the x11 title at 4k F 1 2018 places the RT x26 TF e at 54 FPS average just ahead of the vega 56 red dragon and the gtx 980ti hybrid from two generations ago the overclocked r-tx 2060 with a plus 175 core offset ends up about tied with the stock RT x 2070 but note that this core offset is difficult to sustain without a lot of testing and good cooling the 27 t of course outpaces the 2060 when overclocked maintaining a lead at 68 FPS average and of 14% when both are overclocked versus overclocked versus the stock gtx 1060 from the last generation the 2060 stock card maintains a 59 percent improvement the gtx 960 isn't on this chart due to limited ability to run 4k and the RX 590 ends up at 38 FPS average nearer the gtx 1070 than the 2060 at 1440p the r-tx 2060 ends up at 93 FPS average stock posting excellent 1440p performance and relatively playable 4k performance which is important to note but between the two 1440p of course looks better this is similar to the gtx 1070 TI the to being imperceptibly different and just behind vega 64 versus the gtx 1060 the 2060 posts a gain of 62% which is more than what we saw at 4k the 2060 leads the 960 SSC by 123 percent finally an overclock on the 2060 gets it to nearing error margins of the RT x 2070 stock cards 102 FPS average running 1080p instead we need to pay attention to the scaling gaps between the X X 60 generation GPUs the RT X 2060 place is at 120 FPS average which the GTX 960 gaming addicts by 61% within margin of error of our last distance measured the 2060 leads the 960 by 167 percent down marginally from the lead at a higher resolution for reference making 56 runs 115 FPS average which is imperceptibly different from the 1070 TI and nearing 2060 performance the RX 590 is far behind at 82 FPS average far cry 5 uses the Ubisoft dunia engine and as our next benchmark stressing the geometry pipeline and putting load on the cards by way of screen space reflections and geometric complexity with longer draw distances running 4k the RT X 2060 ends up at 41 FPS average dropping settings would allow this to be a bit more playable of course and increased frame rate but it may make more sense to instead go down to 1440p either way scaling positions this is again functionally equivalent to the 1070 TI it's within margin of error and Vega 56 the 2070 is ahead by a bit but both cards are having trouble with this workload at 4k 1440p shows the same positioning 20 60 places with an error of the 1070 ti it's led by Vega 64 but is now led more significantly by the 2070 which holds a lead of 12% overclocked in the 2060 pushes it to 82 FPS average are roughly tied to the stock 2070 Vega 56 is outside of perceptible differences versus the 2060 with the RX v 9d trailing notably behind at 55 FPS average for perspective the gtx 960 is at the bottom of the chart at 1080p the 2060 ends up over 100 FPS average when stock falling between the 1070 Ti and Vega 64 cards not distant from the gtx 1080 FTW generational improvement versus the stock 10 60s 73 FPS average post a 47% lead not as big as we've seen elsewhere but still a tremendous leap the gtx 960 SSC operated 44 FPS average in the same test demonstrating how far we've come by allowing the 2060 to poll a 144 percent lead again that's not quite as large a lead as an other tests and helps illustrate that the 2060s biggest improvements over maxwell would be in high resolution performance shadow of the Tomb Raider uses a modified crystal engine and should eventually get some r-tx implementation but for today we're testing just with high settings and TAA across all these resolutions 1080 1440 and 4k we just say well they a and some SLI testing but those aren't relevant for this comparison the rgx 2060 ends up at 37 FPS average and 4k testing between vega 56 and the GTX 980 overclocked into 2060 pushes it to 40 FPS average so it's still struggling we'll need more than just a settings drop to get reasonable framerate out of this game as evidenced by the 20 70s baseline performance of 41 FPS average right near the overclocked to 2060 let's drop to 1440p at 1440p the r-tx 2060 runs a 68 FPS average now playable and maintains a 1% in 0.1 percent low performance also nearing 60 this kind of frame time consistency tightness is indicative of an overall smooth frame rate with no stutters interval to interval the 2060 leads the 1066 gigabytes 41 FPS average by 65 percent further leading the 960 SSC by 1 so d 9 percent consistent with previous games the RX 5 90 runs closer to the gtx 1070 with the RT x 20 70 stock card being tied by the overclocked RTS 2060 this makes a compelling argument for buying a cheaper 2060 an overclocked unit in order to achieve 2070 performance at least in rasterization just for reference in this chart we also included the XC ultra 2070 data we'd been pulling it for chart length reasons and the previous ones but it clearly illustrates that higher in 2070 s will outperform the cheaper model and if we technically have a bit more data here we have GTA data we have some or 1080p data but we'll leave that to the article and link it below if you want that and instead move on to something different which is battlefield 5 RT X testing this will teach us something new whereas GTA is mostly a repeat of the previous benchmarks in terms of scaling with battlefield 5 the RT X 2060 tested in single-player with DX are off on low and on ultra is what you're looking at now no this is this is after the recent performance improving patch so BF 5 is at peak performance here with DX are fully disable to be measured at 104 FPS average using ultra settings across multiple test passes with a frame buffering disabled and all post-processing in the basic tab disabled as well like chromatic aberration or lens distortion frame time is posted 61 FPS 1% and 56 FPS gyro 407th which is actually not the best we looked at frame time data and it oscillates a bit between eight milliseconds and twelve milliseconds now that gap isn't big enough to really see or observe as a human but it's getting there it's about 812 milliseconds of the perceptual range for most humans looking at frame to frame intervals anyway we'll look into that more later perhaps after CES but it wasn't the most consistent run we've seen yet and that's why there's loes or not as high as the average DXR on low puts us at 66 FPS average with DXR on ultra drooping to 55 FPS allowing 0.1% to hit 33 FPS average so we'll give the 2060 credit where it's due it's able to play this game at 1080p and roughly 60 FPS with ray-tracing fully enabled or well partially enabled anyway it does lose about 40 percent of its performance in doing so but it can do it so it'd be a question of do you want to play at 1440p Ultra with no rate racing or 1080p ultra no ray-tracing at 100-plus fps or do you want to tank your FPS go down to 1080p and take the ray-tracing and it's sort of subjective but as people who have seen all these options in person we can tell you that generally speaking killing the rate racing on this card and instead going for the higher frame rate in this particular game does feel a bit better but it's going to detent depend on the game like if it's a more in-depth RPG or something where you want that immersion so we'll talk about that if it ever happens let's get into thermals for this we disassembled the r-tx 20 60 and attached thermocouples to the PCB components we have some b-roll of the disassembly process that we can show it's an insane amount of effort to take this card apart and videos use of hot glue to hold in fan cables and seven screws to hold in that set that same fan cable really don't make sense it's just it's crazy what they did to assemble this thing but the rest of the heat sink fan assembly is equally crazy and you can watch our teardown video coming up shortly for that if you haven't already seen it the PCB is the same as an RT X 27 T PCB it's just the cooler has been actually changed in the ecent amount the PCB has been simplified though some they've removed memory modules of course but we already know the hot spot MOSFETs from the 27 t PCB so thermocouples on those starting first with fur mark for torture thermals we measured the stock GPU temperature as 68 degrees Celsius throughout the test ambient was 22 degrees plus or minus one for our regulars this isn't presented this time as delta T over ambient it's just 68 degrees straight setting the fan noise to 40 DBA to normalized noise levels with our other tests we see GB temperature fall to 64 degrees because Nvidia's default fan profile leans up more toward quiet than cool and so 40 DBA a reasonable noise level actually brains a reduction in thermals which isn't all that common MOSFET temperatures end up at 78 degrees Celsius for the stock card which is within spec and reasonable or 73 degrees for the 40 dB a variant on testing these mosfet thermals are within spec and doing fine overall there's really no problem here not until you're 90 plus degree isn't even that it's not really catastrophic we won't bother plotting this but for the curious the auto fan speed it was 1700 rpm or about 45 percent forty DBA fan speeds are about 52 percent for the next chart we're looking at 3dmark fire strike extreme performance over a prolonged period the point is to see how the core clock behaves when under a gaming workload as some clocks will bounce around more heavily based on load first plotting the temperature we see GPU core temperature at 69 to 70 degrees on the Left axis with MOSFET temperature at 76 degrees also on the left axis the frequency cue line comes in next starting initially at nineteen thirty-five megahertz and remember this is stock so it's a pretty high frequency and dropping almost immediately to eighteen seventy-five megahertz once that steady state frequency sustains at this level in a nearly perfect line which is rare and good the drop from nineteen thirty-five megahertz is because GB core temperature steadily rises from 30 degrees up to where it lands driving down frequency pursuant to boost behavior every couple of degrees it matters for frequency overall though and video is doing better on this card than its previous designs and that's because the TDP is lower overall for the GPU core than the twenty eighty series now TDP is not lower in fact it's higher versus the ten series but that's not really what we're looking at here we'll look at that moment as for noise levels we're starting at a lower baseline of 32 percent or 1190 RPM with our TX 20 60 whereas some of the previous our TX cards had a floor of 41 percent or 1525 rpm the our TX 2060 has the same acoustic performance as the other our TX coolers more or less but does plot idle noise lower @ 31.4 DBA the average load fan speed is about 1670 rpm and 22 degree ambient testing environments putting fan speeds at around 37 DBA you can expect it to be a bit higher maybe 39 to 40 in a hot case 100% loads put the our TX 2060 at fifty eight point seven DBA it's still louder than a good aftermarket cooler but it's similar to low-end coolers like the RT X 2070 black by EVGA so this is just going to depend on how much money you spend on the card remember that decibels are logarithmic so these increases are non-linear it's it's not the same scale as you see with other charts people tend to start noticing noise changes for reference with roughly every three DBA increase where there abouts it's a bit of a human perceptual difference there so that's kind of up to you finally from power consumption we're looking at total system power draw with a heavily controlled test bench all voltages are controlled on the motherboard ensuring no fluctuations that would throw off measurements this is not individual card draw but an overtime plot of total system draw first the our TX 2060 ends up at about 260 watts to 300 watts total system power consumption occasionally spiking to 310 watts the average sustained high is about 300 watts flat overclocked in the GPU pushes the system to 360 peak but the average sustained high for the overclocked variants is closer to 320 watts with that OC we applied the gtx 1060 game index with an overclock which is about 240 watts total system power draw for perspective so NVIDIA has gone up generationally alongside its performance in Greece it makes the performance gains a little bit less impressive when you put into perspective the required additional power to do it now it's not a lot of additional power compared to some other cards we've seen in the market but it is an increase and that takes away some of the just how special it was to see there's 50 plus percent gains because it's not really that directly comparable when you have higher power consumption TDP was previously 120 watts and is now 160 watts by Nvidia spec and power consumption difference is noticeable here this is still below the RX 5 90s total system power consumption of 320 to 330 watts but not by much we won't plot it since it make the chart too busy but note also that the our TX 2070 is roughly the same as the 2060 except it's about 10 to 20 watts higher on average maybe lining up at about 15 watts over the whole test so overall the 2060 is very close to the 2070 in power consumption which makes sense given the performance well put over clocks up in on the screen briefly this is the chart of our overclocked progression it's about the same as the previous cards overclocking did pretty well on our 2016 this will be your mileage may vary scenarios where I mean if it's the coolers not as good or if you just got a worse bin then not much you can do about it but that's what we got so it did overclock well the biggest thing here with overclocking is if you like the 2070 performance but you don't like the price because you just don't have enough money to justify it then the 2060 can basically become a 20 70 it doesn't scale linearly in every application but it's pretty darn close and on average 2060 overclocked is about 2070 baseline performance for a $500 2070 and of course you can overclock of 20 72 but if you just want 20 70 baseline you can buy a 20 60 overclock it and pretty much get it you'll be within a couple of percent and it's a great way to save 150 or more dollars so that's something to keep in mind the RT performance might not scale quite the same way we haven't looked into that just yet but we do have some baseline numbers to give you the rest of it now as for the rest 2060 is a strong 1440p performer and of course has no problem with 1080p 4k is is difficult to do with those ultra and very high settings but can be done with medium or lower settings it's just kind of a question of what do you want you know at the frame rate and the higher quality or do you want resolution and on average I think most people generally would prefer the higher quality to just straight high resolution because you'll start noticing the low quality at some point so 1440p for sure no problem 4k if the game is lightweight enough yeah but probably not on average we wouldn't buy it for 4k as a as as the main goal of the GPU versus the RX 590 the RX 590 makes a lot of sense at two forty to fifty to sixty dollars and especially versus the ten sixty although the ten sixty does have a few instances where it's better so if you're only playing GTA 5 just get the ten sixty but that game is pretty old now people do play more than one thing the 590 on average is better overall the 590 versus the 2060 doesn't necessarily make sense as a discussion point because it's a hundred dollar price difference the 590 if you really struggle to justify an extra hundred bucks the 590 is fine it does 1080 it does 1440 mostly fine you can tweak the settings you'll be okay the 2060 obviously different category of price for performance and value it is significantly improved over the previous Nvidia r-tx launches Nvidia is feeling much more confident about this product in speaking with them than the inter market perception than the previous launches the 2080 as we all notice to recap it really had no value over the 1080i it was RT cores and no games at r-tx when it launched or for the 52 days following and beyond that performance is about 1080i performance so there was there's no value there the 20 ATT I had value because it's the best but that only gets you so far when it's 1200 plus dollars so those were much more rocky there are difficult launches friend video the 27 he started to turn things around a little bit the 20 $7,500 cards were decent enough value where you could definitely justify it or argue it over competition including primarily competition for many videos previous generation but once he got past 500 bucks it was a difficult argument actually especially at 550 the value just was was not that great so the 2060 furthers that and does correct things a bit at 350 the value is pretty good overall we would say if you are looking for something that's strictly enthusiast level for fun overclocking we would still recommend Vega 56 it is it's the most tweakable it's pretty open you can do power plate tables mods that you can't do it under the afford any modifications of power limits and v bios still pretty locked down but there are things you can do to get around stuff on AMD words you can't really with anybody you're pretty locked in so if you want just enthusiast level fun genuinely just overclocking for fun then vega 56 gets our vote still if you're going for performance out of the box or you kind of you might overclock for an hour or two and then you never touch it again for three years the 2060 is a good option to consider it outperforms the cards that are cheaper than it pretty much well significantly in most instances significant out performance versus a ten sixty bit more power consumption about 170 percent foreign so up lifts your 2x plus your well yeah hundred seventy percent more performance than gtx 960 SSC for gigabyte card so if you're nine hundred generation that's a significant improvement seven hundred generation significant improvement versus RX 590 it's also holding a pretty big lead on average so it's really vega 56 64 you're kind of looking at and there was more or less flanked the 2060 it so it comes down to price and if you can get a twenty 64 350 bucks it is a pretty easy argument that that would be a reasonable purchase or one of the better purchases overclocking enthusiasts options notwithstanding where we'd still push you towards the vega 56 because that's just it's a lot of fun you can really unlock it do a lot with it but it doesn't mean it's the best performer or the best power consumption especially when it's overclocked so lots of stuff for you to consider on the market these days that's why these conclusions are getting longer and it's more like for option a consider these two option be consider these and it's not just a flat-out by acts by why because it is a complex market so 2016 for out-of-the-box performance $350 we're fine with it it is significantly improved over the previous r-tx launches so good for nvidia there and it can actually somewhat surprisingly do r-tx now that said battlefield fives rate rates and implementation is pretty gross it's really not well done we have a whole video I'm not going to go into it here you can just search the channel for like battlefield 5 RT X and our in-house 3d artist an expert and all this stuff helped me analyze it and did a great job where we talked about the shortcomings of RT acts and the good things that it does so that's really not a compelling reason to buy the 2060 the performance is though and the rasterization performance I will say the founders Edition cooler is completely insane it makes absolutely no sense they made it worse than the 20 80 and 20 80 TI it's using now it's still using glue but now in more places it uses glue to hold the cable in place uses seven screws to hold the fan cable in place they've got soldered connectors wires for the PCIe connector which are a a risk of being broken if you need to service the fans or replace the paste for very common things to do overall and they changed the screw sizes from the previous ones they still have just as many screws really the the if Nvidia ever gets these back if you have one the fan dies you send it back they're gonna throw it away there's no way it's worth the service technicians time that they're paying him or her to repair a card that you can barely even open so it's just it is functionally impossible for a novice to service and it is difficult for an expert to service but if you want to buy something for a challenge I guess by 2060 and try to take it apart so that's it for this one thanks for watching subscribe for more his always go destroyed I came is Nexus net to pick up something like one of our brand-new copper Mule mugs with the thermal conductivity of copper written on it because we think we're clever like that where you go to patreon.com/scishow cameras access it directly stay tuned for CES thanks for watching I'll see you all next time