Gadgetory

ever since getting the studio space we finally had some time and excuses to do content that we expressed interest in doing about a year and a half ago so this is the r9 290x you might remember it it's one of the previous this specific one is a reference card from AMD from several years ago predated the 390x predictably the RX 480 the fury X the rx 580 and so forth so it's a good couple generations old depending on how you define generation this this this instance and today we are retesting it for modern games so we're looking at DirectX 12 API is looking at modern dx11 implementations versus existing AMD and NVIDIA products to see how it stacks up in modern gaming and if it's worth at this point upgrading before that this video is brought to you by mass drop and the PC 37 X gaming headset with professional-grade Sennheiser noise cancelling microphone the PC 37 X headsets are what we use in the office for phone interviews where audio and mic quality are critical making it very convenient high performance solution for gaming or professional work the headphones come with a detachable ten-foot cable for safe storage during travel a standard 3.5 millimeter plug and soft foam 4 firm but comfortable fit over the ears learn more at the link in the description below so like I said we wanted to do this content a long time ago originally just now got to it but we're here now so this is the only 290 X we have still in the lab and it is a reference model but before anyone gets upset that it's not going to be representative of their own cards performance note that we overclocked this to a point where it's it's a bit faster than most of the AIB partner models of the 290 X when they launched so we have stock and overclock numbers the overclocked numbers represent pretty much any overclocked to 90 X because we did max out the fan on this to make sure it could reach those higher frequencies and my hearing has now been permanently damaged as a result of it because it's about 60 decibels when it's running at full speed so try having that next to your head for about 6 hours and other than that I mean the card was a flagship it was high-end and this particular models four gigabytes the 390x popularized an 8 gigabyte alternative that was a refresh with 290 X and 4 gigabytes though before most of these tasks you'll see is actually fine because it becomes a limited elsewhere first for differences versus other cards look at something like an RX 580 which is an RX 480 that card is a modern well it's kind of a flagship it's a bit of a weird place to be because it was a flagship of Polaris but it was never really a high-end product there's always mid-range competitor with something like 1060 but this reverse is that so the 290x has more TM use at 176 verses 144 texture map units on the RX 580 which is a 480 it's got more ROPS at 64 versus 32 now the the capabilities of those ROPS will change over time as architecture advances but having more is still going to be beneficial and this in particular is something that we want you to pay attention to during these benchmarks because with the increase in ROPS there's a benefit to things like anti-aliasing when we do our GTA 5 testing it uses a to tap msaa it also benefits higher resolutions kind of the same thing there were you doing with multiple samples with AAA or it's more pixels with high resolution so ROPS will be where you might see the difference between a 580 and a 290x close as the resolution increases as those Rob's are tapped more heavily in their rendering workload so that's something to pay attention to and Rob's it's it's it's where the output is effectively assembled into what becomes the image a bitmap image at the end of the pipe so we're testing with a reference card we overclocked it blasted the fan speeds all that stuff let's get through the testing today we've got a lot of modern gaming we have GTA 5 is an older representation but this card is circa about two thousand thirteen or fourteen or something so it is quite old at this point but it's worth testing so let's get into it and then we'll talk conclusions Sniper Elite 4 will start us out before displaying results remember that one of the most interesting areas to look is going to be for scaling performance between two fixed goal posts as we change resolutions if the distance between those posts tree so that's indicative of an architectural advantage or deficit at the new resolution will set the RX 580 and r9 290x at stock settings for these goal posts Sniper Elite 4 gives us a well optimized dx12 title to test with which is valuable because we want higher frame rates even at 4k to better illustrate some of those scaling gaps keep in mind that the 290x came out long before 4k was popularized and also before DirectX 12 was even released 1080p still had almost all of the market share at this point at 4k we see the r9 290x stock card at 38 FPS average with it lows it's surprisingly close by at 32 and 30fps 1% and 0.1% low well look at frame times in a moment but just stick with the averages for now overclocking Headroom was limited and capped at about 10 60 megahertz for core getting us to 41 FPS average and climbing 7.9 percent over the r9 290x card these performance figures pegged the 290x and its overclocked counterpart at rough equivalence with the RX 588 gigabyte card not too distant from the new rx 590 Fatboy this is without yet considering power consumption mind you so that field of things has potentially changed a bit as well the GTX 1060 is just surpassed by our 290x results as is the GTX 970 as for those posts we mentioned the r9 290x stock GPU allows the RX 588 gigabyte stocked review to hold the lead of about 2.9 percent transitioning to the more limited 1080p results we now see that this 290x has a stock frame rate of about 98 FPS average allowing the RX 588 gigabyte card a lead of 9.4% with its 107 FPS average the fact that the r9 290x closed the gap at 4k suggests to us that the 580 becomes limited and its ROPS and TM use but primarily ROPS the 290x is better equipped on this front just in pure Roffe account leaving its biggest limitation as frequency which is why the card has more trouble keeping up at the lower resolutions once again the orange 580 gets slammed with higher demand on the pixel pipeline where comes more limited the r9 290x pulls ahead and that's kind of where we see the shrinking gap the same pattern would emerge with anti-aliasing as well as it's effectively increasing the sample rate in the same way or at least with the same intensity that increasing resolution would do this is also becoming Rob's bound at least when you start modifying the anti-aliasing settings so even if you play with lower resolutions on 4k but you like higher AAA or something like that you'll still see these same types of behaviors emerge as for 1080p performance on the whole the 290x performs behind the gtx 1066 gigabyte and ahead of the RX 574 gigabyte cards frame times are what we're most curious about as a reminder frame time plots demonstrate the frame to frame variants and time to present a new frame this is a measure of frame to frame intervals in milliseconds so lower is better versus benchmark progression on the bottom axis the more consistent each point on the line is to the previous the better the user experience deviation from the mean and access of 8 to 12 milliseconds becomes noticeable to most gamers the r9 290x does well in this department with sniper elite for its a sniper's credit the game is remarkably well built but the 290x still needs the right hardware to keep frame pacing consistent in this title we don't see too much deviation from the mean frame time with the biggest variants at least frame to frame variants in the form of three to four millisecond swings this is completely acceptable and as you can see isn't too distant from the modern rx5 90s performance the 590 has fewer peaks on average but the difference in consistency is unnoticeable overall for most players the r-tx 2060 is also plotted as an example of the most modern architecture where we're nearing an ideal frame time plot the takeaway though is that the 290x does well on frames I'm consistency in this particular title and that's despite some early life issues with frame time consistency in the 290x many of these were patched up with later driver updates but the rest would likely be more game or API dependent f1 2018 is next giving us a DirectX 11 game that uses the same API as almost all of the market back when the 290x least gesture scaling reasons we'll look at 4k results despite this card not really being meant for it and especially 2019 at 4k the r9 290x for gigabyte card ends up at about 33 FPS average right here that's similar to the gtx 1066 gigabytes and gtx 970 the RX 588 gigabyte outperforms the 290 x 4 gigabyte by 4.6 percent landing at 34 FPS average we can also learn from the 390x results which shows a 34 FPS average this card is a refresh of the 290 X with a higher frequency and double the memory capacity in this title it rapidly becomes clear that the memory is not the primary limitation as performance only increases by a few percentage points and the loads don't really change all that much the 290x and 390x are more limited by the GPU than by the memory moving on to 1440p we see similar resolution scaling as the previous game the RX 580 stock GP is 56 point 6 at PS average is nine point one percent ahead of the r9 290x is 51.9 FPS average protein a relative game and performance for the RX 580 again we think this is because the 290x can leverage its increased ROPS and texture units at higher resolutions or higher anti-aliasing values closing the gap as resolution increases that doesn't mean it's playable out those higher resolutions or that is better just that the gap closes it illustrates how the GPUs of scale for 1440p resolution the 290x is still reasonably playable in this title dropping settings from ultra high to just high or similar would make for consistent 60 FPS and Beyond comparatively the 290x does about as well as the gtx 970 although the 290 X's at lows manage higher results with the 390x not too distant the 390 X's extra memory doesn't get leveraged in a meaningful way for this benchmark versus some modern cards the 290 X is outperformed by the GTX 1060 and rx 580 alike we don't see too much improvement for the 290x at 1080p moving up to 65 FPS average and with still minimal gains from overclocking so rx 588 gigabyte runs at 72 FPS average for a lead of 11.6 percent to recap this title we see 11.6 percent improvement in the RX 580 at 1080p and one percent at 1440p and 4.6 at 4k showing very clear performance improvements in the higher frequency newer cards particularly at lower resolutions shadow of the Tomb Raider is up next giving us a DirectX 12 title for another modern look at performance DX 12 didn't officially launch until 2015 so the 290x was made well before the new API saw any adoption at 4k the 290x obviously struggles at 24 FPS average making it largely unplayable with these settings the rx 580 doesn't do much better at 25 FPS average with the rx of 5 at 90 at 28 FPS average let's move on to something more reasonable have 1440p the 290x runs at about 42 FPS average with the gtx 1060 functionally tied with the 290x the differences are inside of error margins so we can't state if one is better than the other in any meaningful capacity the rx 588 gigabyte leads at 45 FPS average with the 590 at 50 FPS average and Vega 56 at 62 FPS average 1080p positions at the r9 290x in playable territory even at with these higher settings at 60 FPS average for the overclocked version and that's about where most partner cards would fall in 58 FPS average for the stock model that puts the 290x as comparable to the gtx 1066 gigabyte and behind the RX 588 gigabyte far cry 5 uses geometrically complex meshes and longer view distances making it one of the more draw call intensive benchmarks we run for this at 4k far cry 5 positions the 290x at 26 FPS average right between the gtx 970 and 980 cards and 40 in the RX 580 a lead of 8.3 percent at 28 FPS average this isn't particularly playable under these settings so once again we're mostly using them for perspective at 1440p performance climbs significantly it's a 43 FPS average which is about where the RX 570 and 390x perform and videos gtx 1066 gigabyte out performs at the 290x by a few percentage points here with the most modern cards posting significant leads the RX 580 holds a 16.7% Lee showing one of the largest gaffes we've seen between the two yet but still following the trends that we saw previously 1080p really carries this trend though now allowing the RX 588 gigabyte car at a lead of 22% which is the biggest gain we've seen last far as a reminder that's against 8.3% at 4k and 16 white 7% at 1440p so our earlier Theory remains consistent as for raw frame rates the 290x is still adequate for 60fps and far cry 5 at 1080p with these settings but it's getting a bit lon on the tooth Vega 56 and GT x10 70s or even RDX 20 60s would offer considerable performance improvements as you can see in the chart is just a question of do you have budget or do you really care because if you're happy at 60fps 1080p the 290x is still holding on reasonably in most of these games GTA 5 is a 2015 game and is the oldest on our benchmark but it's also the most played game out of everything in the test suite at 4k GTA 5 lands the r9 290x at 2004 FPS average which is within error of the RX 580 or even slightly leading it for the first time all bench and not too distant from the GTX 970 this is more of a synthetic look of course since it's not particularly playable 1440p again post the r9 290x and RS 580 as roughly equal in performance we run GTA 5 with 2x MSAA so it's likely we're seeing a potential ROPS limitation still on the RX 580 it's a good example of how anti-aliasing also impacts performance with these cards the 390x does actually post meaningful improvement over the r9 290x here and in at 53 FPS average but this is clearly more of a change in frequency than memory capacity as the overclocked 290x isn't too distant from the 390x stock card so the cards really not all that different from some of the mid-range modern cards that power consumption it's a bit of a different story there but and also the thermal design of this cooler was terrible but if you had a partner model then that is less of an issue or a non-issue so it really just comes down to the performance and of course if you have a 290x yes there's really no harm you feel like upgrading now you can buy a modern 300 to 400 dollar card and get a pretty substantial uplift like pagan 56 for example in these charts you saw a vega 56 is significantly improved over the 290 X and that's further and if you feel like going through some of the power mods that we did with the registry with power play tables so plenty of options there the the RT X 20 series 2060 is an upgrade and it is better it feels a bit odd because I guess in terms of spend you potentially add a bit less than you were when you bought this originally depending on what price you bought this for so 2070 might be a more more comparable price the original purchasing point of a 290x but either way yes there are significant improvements of course in the market that's really not unexpected but the interesting takeaway is that the 290x does better at these higher resolutions than relative to other cards that are on market today then at the lower resolutions and that's because of its it's beefed up ROPS pipeline and extra TM use and things like that the memory capacity didn't seem to affect performance all that much is that unfortunately you can't really leverage the advantages and things like 4k or 1440p often anyway 1440 bit more because the card just starts falling behind I'll swear like frequency overall where Anthea has seen significant improvements over the last couple of years so should you upgrade well it's entirely up to you you can decide that but hopefully you've got at least some understanding now with the charts of where this would fall versus today's card so when making a purchasing decision for an upgrade you don't actually buy a lateral move and get something that's equivalent like rx 580 it's a bit better far cry 5 it sees a significant improvement at lower resolutions but overall @rx 580 is not really much of an upgrade neither is a GTX 1060 for example you want to go a bit higher than those cards so that said for a once flagship to equate the kind of mid-range cards of today it's not bad for the mid-range cards of today it does show progress and especially progress in price that's a good thing but the goal posts always move with games because games just get harder to render anyway so that's it for this one though this was requested by a lot of you after I'd originally suggested the idea and then we forgot so here it is thanks for watching subscribe for more go to store it I came in Texas that's not directly or patreon.com/scishow gamers Nexus I'll see you all next time