Gadgetory

welcome back to harem box for more dirt for action over the weekend I had a quick look at what it takes to play codemasters newly released racing title and in the end found gamers really only require fairly modest hardware to enjoy the game at all of its glory at 1080p targeting 60fps I also got a bit sick over the weekend just to come on cold so nothing to worry about I didn't let it slow me down too much so there's plenty of benchmarks the over but if I stand a bit on on this one more than usual that's what I'm blaming it on anyway intel's pentium g 45 16 am neither radeon RX 570 proved to be the perfect pairing despite recent driver updates from both AMD and NVIDIA on the 9th the red team still looks to have the edge in this title and we'll look at the numbers shortly AMD's crimson Reliv edition seventeen point six point one driver cleaned up to a 30% performance improvement over the previous driver that's obviously a massive difference and that claim was made when using the RX 580 with eight times MSAA enables however the ultra quality preset which is the highest preset in the game and the highest level of quality things i will be testing with only enables for time MSAA but you can probably still expect optimal performance across the board with this driver as for NVIDIA there 380 2.53 WH GL driver claims to provide an optimal gaming experience for dirt for that said given what you're about to see I'd say there's already a more optimal driver in the works for this video I will be focusing exclusively on current and previous generation GPUs and testing takes place at 1080p 1440p and 4k please note the rx 480 and 470 have been dropped from the results for this one since they are very similar to the newly released rx 580 and 570 without them I'm still tested a grand total of 27 GPUs so I feel like there should be enough there for a solid reference point all testing has been conducted using my core i7 7700 K test system clocked at 4.9 gigahertz and I won't be including any other cpu results for this title do it Ford doesn't require a powerful CPU which is why I went with the G 4560 previously the game mainly heavily loads a single thread and that most uses two threads efficiently beyond that on most modern quad cores or greater you will only see 20% or less utilization on additional threads that doesn't have amine the games call the optimized it just means it's not CPU intensive this isn't an ARMA 3 type scenario for example or a single thread gets pounded and the CPU is still the performance limiting factor in other words I'm not seeing a seafood bottleneck in this game of course it would be better if the game engine distributed load more evenly but for a game like don't fall that's primarily GPU bound it's not an issue for those of you wondering I saw almost no difference in performance when using the GTX 1080 TI @ 1080p using either the pentium g 45 60 or a heavily overclocked core i7 7700 k the cpu really does make little difference here horizon v 1400 1600 and 1700 for example also delivered very similar performance and again we're able to roughly max out the GTX 1080 Ti using an NVIDIA GeForce 10 series GPU the game consumes almost 4 gigabytes of vram at 1080p using the ultra quality settings about four and a half gigabytes at 1440p and around six gigabytes at 4k using AMD's Radeon rx 588 gigabyte these figures were slightly inflated hitting four and a half gigabytes at 1080p just over five gigabytes at 1440p and around seven and a half gigabytes at 4k so this again is evidence that in videos memory compression technology is slightly superior when it comes to RAM usage that is to say system memory the game typically consumes around four to five gigabytes depending on configuration the exception to this will be seen when using a graphics card that features a limited via Ram buffer at resolutions that can't handle anyway for the most part eight gigabytes of RAM should have you covered for this title okay well I think that's about everything let's check out how the various graphics cards got on so kick starting things off we have the historic lineup those big old 28 nanometer GPUs from a time a long long ago 2014 I think it was anyway despite their advanced age the B team does a great job in this title at least the mid-range enough does anyway right away we see the gtx 960 trailing the r9 380 by 17% margin while the GTX 970 was 31% slaw from the r9 390 in fact while the gtx 980ti was able to pull the head of the r9 390 by a small margin when comparing the average framerate it was 5% slower when comparing the minimum therefore it was the fury ax that pulled ahead at 1080p rendering an impressive 96 fps making it 20% faster from the 980 TI on average and 31% fast when comparing the minimum frame rate the higher-end NVIDIA GPUs are suffering some kind of issue when it comes to the minimum frame rate which is very strange and well worth keeping an eye on upping the resolution to 1440p changes things quite drastically and now the NVIDIA GPUs come right back into play it's almost like we were seeing a driver overhead issue for NVIDIA at 1080p wouldn't that be ironic anyway at 1440p the GTX 960 is now just a single frame slower than the r9 380 that said the GTX 970 is still quite a bit slower than the r9 390 trailing by 25% margin the gtx 980ti is able to close it on the fury x though it is still quite a bit slower when comparing the minimum frame rate in fact the 46 FPS minimum of the 980 TI means it can only match the Nano and r9 390x now the extreme 4k resolution none of the GPUs tested were really able to deliver satisfactory performance so the fact that in video has caught up here is of no consequence the fury X managed 36 fps while the gtx 980ti averaged 37 fps and with minimums around 30 FPS the input lag was quite difficult to deal with ok so now it's time for the fresh new silicon to show what it's got previously 96 FPS at 1080p was as good as it got with the fury X and I have to say that result looks even more impressive as the GTX 1080 pushed just 106 FPS on average yes the Titan XP and GTX 1080i were faster again but naturally you would expect them to be further down the food chain we find the GeForce GTX 770 and things aren't all what you would expect with an average of 79 FPS the 1076 alongside the RX 580 in fact it's one FBS slower when comparing the minimum frame rate meanwhile the gtx 1066 gear bar is just 30% slower than the RX 580 when comparing the minimum frame rate in fact even the RX 5 70 beats at the gtx 1060 so again more evidence than has some work to do jumping up to 1440 pcs very few GPUs capable of achieving 60fps though once again the margin between AMD and NVIDIA does close up the GeForce GTX 1060 isn't a great deal slower than the RX 580 now or at least compared to what we saw at 1080p the gtx 1070 pulls ahead of the RX 580 providing slightly better minimum frame rate performance with a much stronger average then finally at 4k it's just the gtx 1080i that breaks the 60 FPS barrier I shouldn't know that the gigabyte or SD 2 X 1080 I was used for testing while the Titan XP is obviously a reference card now throwing all 27 GPUs into a single graph we get this mess here we get a sense of just how well the fury X is performing in relation to NVIDIA GPUs right now it was just 9 percent slower than the GTX 10 a DTI and Titan XP when comparing the minimum frame rate at 1080p it was also quite interesting to see graphics cards such as the GTX 1060 getting completely mobbed by the previous generation r9 390 the issue seen at 1080p for NVIDIA seem to be resolved at 1440p as now the GTX 1080 is significantly fast in the fury X even when looking at the minimum frame rate that said the GTX 1070 still loses out to the Nano and 390 X for the minimum frame rate then at 4k we see it really takes a current generation GPU and even within video not hitting full stride at least in my opinion they still led the high-end benchmarks by a country mile last up we have some quality preset testing which has been carried out at the 1440p resolution using the gtx 960 and RX 580 some interesting results can be seen here so all the results seemed previously were based on the ultra preset and here we see the RX 5 that is almost 30% faster oh and please note that for the preset testing I have used a different section of the game for testing so the numbers here won't align with those seen previously though the performance trend will be similar anyway whereas the GTX 1060 was almost 30% slower using the ultra preset it was just 11 percent slower using high but as we move to medium the 1060 pulls ahead is now 10% faster and this was also seen when using the low and ultra low presets as well there is clearly something enabled or turned on with hi Ultra presets that hurt in video more than AMD while it's great the GeForce owners can boost performance dramatically by lowering the quality settings it's all AMD right now where it counts before wrapping things up here's a quick look at the various quality presets side-by-side recorded using a Titan XP at 4k that's six of a crash a break for 60 right three I'm pay a bit left and break6 let's six left six sixty right five of above 80 left six of them 16 right six in the best one law right now AMD is showing very strong performance in dirt for particularly when looking at the minimum frame race that said if you back off the quality preset lowering from say old shredder high nvidia does start to catch up the gtx 1060 for example so a massive sixty percent boost in performance while the RX 580 improved by 40 percent and I suppose either way that's some pretty big numbers for both AMD and NVIDIA for what was a very minor downgrade in the visual quality running the game at high allowed both graphics cards to average well over 60 FPS at 1440p so unless you're running extreme hardware I highly recommend using the high quality preset in favor of ultra and of course this is pretty much true for most modern games as usual the vegetation completely hammers performance when set to ultra circuits that will light on vegetation so desert circuits I would run it over 70 FPS on a mid-range graphics card then those with heavy vegetation like where I tested that figure was pretty much hard again when it comes to CPU performance the game really only uses one or two threads and while that certainly sounds bad it's not really as this game just doesn't require a cool app you see for you to perform well and we found this over the weekend with the pentium g 45 60 as it was able to provide a smooth and enjoyable experience in dirt for well that brings me to the end of my dirt for performance coverage for now I'll be sure to keep an eye on driver developments and report back with what I find I'm your host Steve see again soon guys you