Terrible Optimization: Cities Skylines 2 GPU Benchmarks & Graphics Optimization Guide

[Music] City skylines 2 is one of the worst optimized games we have ever tested and normally we don't throw around the word optimized but with this one it's really clear in fact it was bad enough that the developers themselves posted multiple panicked updates ahead of launch to tell people as much in the time since launch we've tested about 23 gpus at multiple Graphics settings and resolutions but we won't bother presenting anything beyond 1080p because the RTX 4090 was stuck below 40 FPS average when it was at 4K low city builders are actually one of the main genres that I've played over the years so I have a lot of experience dating back to Caesar 3 was the first one I ever played I love city-builder games this one it just has a lot of performance issues we're going to talk about some of those today and uh this tasting was done using a pack of save games that were published by city planner plays who's a YouTuber who really focuses on the gaming side of this so city planner did a good job with some of the initial benchmarks we're here to pick up the torch for some in-depth performance and Graphics analysis where we found everything from bugged texture quality settings that almost seem inverse to the setting all the way up to bad stutters the game itself appears to be getting updates so that's great to see from what we can tell it looks like a fun game that just needs some more time to bake and the fact that the devs can increase performance within 2 days of launch proves that but optimization or not we can still talk about GP performance so let's get into that before that this video is brought to you by Lexar Thor memory lexar's new Thor ddr5 and ddr4 RAM includes a heavyduty aluminum heat spreader while retaining a relatively low profile form factor for PCs with tighter air cooler clearance the Lexar Thor memory supports both Expo and Intel XMP timings and is available from ddr5 5600 to ddr5 6000 cl32 learn more at the link in the description below there's a lot in here considering it all took place within the last couple of days for us so quick breakdown of what this content is we have comparative benchmarks that means just a lot of cards against each other to help you choose which one you might want if you're buying or upgrading we also have oneoff feature test for optimization purposes where we apply each setting individually to look at the type of impact it has if you're trying to boost the frame rate by a little bit while still keeping as much Graphics quality as you can and additionally we have a section dedicated to the visuals of the game comparatively looking at one preset versus the next just with all the settings changed per the preset to see what you might be giving up by going from say high or medium down to low because the patch shift after we' done our testing with all of these gpus we went back to add in the patch that will give us relative percent uplift scaling from a couple of cards so that we can figure out how much it might affect all of the cards on the chart because we're not going to redo all of them for 5% but we can get an idea where things land now with that current patch most of these cards are still roughly in the same same place that they were with the launch version of the game which is what we did most of our testing with so they're within about 3 to 5 FPS or so of where they were data logged originally and uh that won't change in many cases however it is enough to help in a few specific cases especially cards that are sort of on the edge of playability let's start with researching the city size for defining the methodology we did this testing with our standard GPU test methodology and platform but we used the city planner plays files for testing so again a massive thank thank you to him for posting some real gameplay save files to make testing easier we'll link him below if you play this game you should check out his channel a game like this you can't test properly just on an empty map or a freshly started one most city builders the performance degrades over time as you continue to add on to the city so let's look into that we were able to do like for like benchmarks at different city sizes but with the same Bones the weather and time of day conditions were all the same as was the test course so just the city developing was the change and the impact here is crazy here's the chart this was conducted with the patch released on October 26th so that's the first performance patch this chart shows the 490 at 1080p medium settings and the scaling is clean at 1,000 population pretty close to a blank map we're at 115 FPS average that's insane as compared to our 100K population map where we see a 62% jump if you were to demo your city and evict everyone back down to a population of a thousand to have a game with a 62% range is uncommon we rarely see more than a 10 to 20% swing and like for like testing and games in this game though the longer your city is developed the lower the frame rate will be but the performance degradation did appear to slow and become asymptotic as we approached 100K population the 100K map though was used for our testing as it's a heavier workload while still representing a realistic gaming scenario we'll come back to more research after the comparison charts to show some optimization tips we'll start with the impact from the updates when the update launched we were actually already filming this video but we decided to pause the entire project to investigate the new patch this however was after we tested about 23 gpus in three different resolutions with several different Graphics settings amounting in over 500 test passes so the update included some specific notes on stutters and on LOD at higher resolutions alongside some Global illumination changes we'll talk about those later in the video for now though let's start with the impact from the patch alone to Baseline just showing a small subset of gpus for the percent scaling in this chart you're looking at the original passes versus the patched passes at 1080p low the RTX 490 posted a 5.6% performance uplift from the patch it's not much and we're not at the CPU limit yet that happens closer to 112 FPS the RX 7800 XT posted a 4.5% Improvement in average FPS that's good but from the perspective itive of our original round of tests not much has changed you can approximate these differences in the upcoming charts the RTX 4060 saw an improvement of 99.8% so that's more meaningful than the other two and the RTX 360 posted the most meaningful yet at 11% that moves it from 30 fps to 33 FPS average it's not game-changing but it's definitely better the RX 6600 posted about a 15% uplift from 25 to 29 FPS so that's a large jump percentage wise as well and the gtx970 was unplayable with these settings originally and remains unplayable here the uplift was massive as a percentage it's almost 18% but that just means it moved from 11.7 to 13.8 FPS average and with the standard deviation on the specific card of about 0.4 FPS that's a large part of that number anyway the trend here is that generally speaking the worst the card did the more uplift it got there was more ground to gain even without a CPU bottleneck though the higher performing cards just didn't benefit benit as much from this patch they weren't Bound in the same way so seeing 5% or a little bit over that isn't uncommon on the higher end as for frame times there was an improvement in most of the cards not always but generally speaking it appears to be better uh but there are still massive frame time and frame pacing issues now we're going to run through the original performance numbers from prepatch but keep the numbers we just looked at in mind to help you calibrate against the new version so to be very clear what's coming up is from the day one launch of the game which is when we started testing it uh it'll give you the most gpus to look at but you do need to keep in mind that there's about a 5 to 10% increase on average it is mostly linear and the worse the card did the more uplift there is our first test will be for 1080p medium for our testing we always toggle Dynamic resolution completely off because dynamism for resolution impacts testing and it removes like for like comparisons at 1080P and with the medium preset the RX 7900 XTX sets the C in at about 67 FPS average lows are paced better here than they were at other Graphics configurations with more consistent frame time pacing this is in alignment with the workload going up the 490 is tied with the 7900 XTX here they are indistinguishable and the RX 7900 XT Trails giving the XTX model and the 490 about an 11% Advantage the RTX 480 follows that progressing through the list the 4070 TI ran at 46 FPS average where the 40 80 has a 22% advantage and remember this is 1080p medium these are normally numbers we'd see at much higher resolutions or at least graphic settings after this the game's recommended RTX 380 landed at 40 FPS that's the recommended card that's just below the 6800 XT the low end of the chart is built up by the RTX 460 at 23 FPS average which could be made to work better with lower settings we'll next look at low settings the preset option toggles Dynamic resolution scaling quality to constant again we disable that for testing purposes so we have comparable data the first thing you need to know is that the frame times of this game are completely crazy sometimes this has gotten a little bit better with the first patch but it's still bad in fact here's a look at the RX 7800 XT after the Patch that helped with stutters this is at low settings one of these two passes had several violent stutters over 100 milliseconds including one big spike to 322 milliseconds that's one third of a second that we're staring at the same frame so this is a big problem still the other pass we're showing here is more stable the game can run smoothly but it's not reliably smooth the point is that 1% and 0.1% lows in the charts coming up are not an indicator of card quality this time like they normally are because the game itself is so variable so they're more of an indicator of the game Behavior so here are the results for 1080p low the top of the chart is pinned to around 83 FPS average with the 49 90 and 700 XTX at the front and closely trailed by the RTX 480 if we were to draw a line at around 30 FPS then the RX 6600 and everything below it falls short and would need further settings reductions this includes the RTX 2060 the GTX 1060 and on a technicality the RTX 3060 although 29.9 is effectively 30 FPS average assuming you're okay with that frame rate in this game while these cards were more playable in the lower population cities or the empty Maps they suffered greatly as the population count and the city complexity increased shifting Focus back towards the rest of the chart the skylines 2 Community seems to be regularly talking about vsync on Reddit we ran it on a 4090 just to reiterate an ancient point vsync will cap to the screen refresh rate so on a 60 HZ display that'll be a 60 FPS maximum value that's exactly what we got doing this can help eliminate frame tearing but will introduce potential stutter problems and scenarios where frames are not ready on the flip interval the RTX 380 ran at 61 FPS average here that's the card the developers have as they recommended the GTX 970 was its minimum and that one struggled hard even at 1080p low we could barely exceed 11 FPS average turning depth of field completely off and setting depth of field quality to off just for good measure we saw an increase of about 9% to 12.8 FPS average so percentage wise that's a big jump but it doesn't really help the 7,000 series RX gpus have some interesting behavior in this game they're more advantaged than we often see in some other games so the RX 6800 XT generally outperforms or more closely equates the 7800 XT largely because it has a CU count or a compute unit count Advantage but we're seeing the 7800 XT pull ahead in these charts the same goes for the 6950 XT and the 7900 XT we normally see them closer together or sometimes even flipped depends on the resolution as well very low configures almost everything to a simple state of being toggled off first off we're becoming bound at about 110 FPS average so everything from 104 to 112 on this chart is at least occasionally hitting other limitations on the GPU let's start our focus on the 970 which is definitely GPU constrained we're seeing a result of 28 FPS average maximally a depth of field toggle would gain you 2 to 3 FPS here so that's close to that 10% number again the lows are bad though you're going to encounter stuttering frequently and in particular rapid zooms or camera movements will provoke them especially in the test City that we're using because this game seems best played at 1080p the RTX 3060 TI is actually beaten by the 4060 TI if you've been around for these cards you might recall the memory bandwidth limitations on the 4060 TI were severely hampering it generationally which often results in regressive performance against the 3060 TI when resolution is increased but back to City skylines 2 at 1080p it's ahead amd's newer RX 7600 is roughly tied with the 1080 ti so you'd be able to increase the settings quality on these cards the 3060 is positioned similarly as is the 2070 as for the 40609 ti that leads the closely competing RX 7600 by about 10% the biggest thing we saw here though was better frame time pacing although not great overall most of these cards are at least in the 30s for 0.1% lows our next chart looks at individual graphic settings for purposes of that we wanted to choose a card that was at roughly 30 FPS average when it was at Baseline 1080p low the reason we chose this set of numbers is because the gap between low and very low is gigantic it is one of the largest gaps we've ever seen in presets this low down the stack so we chose the GTX 1080ti ftw3 that we have that meant moving from 32 FPS average on 1080p low to 72 FPS average on very low with the launch version of the game that's massive we normally see swings closer to 10% in games between settings that's a good and a bad thing the good thing is that it means there's a lot of room for players to manually tune the settings where you can make it look better than very low but you can still get some of the performance advantages by getting away from just the low preset so there's room there you just need to do some tweaking of it and that's what we're going to help you with we didn't do this testing for every single card because not necessary what matters here is the relative scaling so we're looking to identify how much performance can you gain by changing different settings down doesn't matter if you don't have a 1080 TI also if we did it for every card it would have been like 322 combinations just for low and very low to be compared so relative scaling it is here's the chart Baseline low with Dynamic resolution off as it was for all these tests is at the top toggling depth of field completely gave us a 33.6 FPS average for both variations that's an improvement of 5.7% on the 970 we saw closer to 10% when we tested earlier but the frame rate was also much lower so the impact may have been disproportionate on that card each entry only changes the setting listed so depth of field remains on physical for the rest turning off volumetric effects got us to 35.2 a boost of about 11% that's significant and could be a place to find a lot of extra performance we also did a test with the Colossal order recommended settings that they published when the game launched so they recommended depth of field off volumetric off and motion blur off there was functionally no impact as compared to the other tests volumetric alone achieved similar results so the gains didn't stack here for average or not by much we did get a stacking effect materializing in the frame time pacing though Shadow quality jaunted Us in huge ways gaining us 37% performance we'll show you what that looks like momentarily but this is huge level of detail scaling at least for the launch day patch boosted us 29% as for ambient occlusion clouds anti-alien in from the FXAA Baseline and texture quality on a card with sufficient vram none of those really affected things much you could leave these at low or boost them higher to increase Fidelity without much loss of frame rate fog boosts frame rate when disabled and since some people seem to really hate how the setting looks anyway that's potentially a win-win Reflections didn't do much in our test environment but might have more impact in other specific scenarios Global illumination was a longside volumetric as having moderate gains time to look at how the first patch affected the game we reran a few of our tests to quickly check on if the scaling we saw on the last chart has changed and that's what matters it's the relative scaling not just the absolute performance here's the chart we reran test for the Baseline low result for depth the field and for Global illumination low Baseline improved to 35.6 from 31.8 and more importantly had a big uptick in the low-end performance we don't have the standard deviation shown here because it's across two different passes so that's a 12% Improvement in just the Baseline the original depth of field toggle was a 33.6 FPS average result which was an uplift of 5.7% on this card from the original low Baseline the patch boosted that to 36.6 so we're only seeing a 3% uplift now from the new Baseline you can't compare the unpatched low base to the patch depth of field toggle that doesn't make any sense the Improvement is less significant now than it was before which suggests changes to depth of field at Baseline are affecting that result the global illumination toggle previously improved performance to 35.5 FPS average which is an uplift of about 12% against Baseline the new result is 38.2 or an uplift of 7.3% against the new Baseline the reduced Improvement again aligns with the Patch's changes to Global illumination this game is performing still poorly overall but the patch brought a huge Improvement they really should have just waited the two days it took to push this let's get into some of the graphics comparisons so because Shadow quality had some of the biggest impact on performance we'll spend some time showing a side by side of the setting this is only comparing the low option versus the toggled off option when Shadows are toggled off you'll notice the 3/4 angle of the city lacks any shadowing at all in the trees in the lower leftand corner toggling between just low and off we have a very flat looking image when taking those Shadows away completely they should just call this option Sim City 3 000 looking at the buildings in the city we can also see a removal of Shadows along the inside walls of the buildings toggling it on and off a few times you can really see the impact of even just going to low Shadow quality but remember on our 1080ti case study from Baseline low the cost was 37% zooming into the city makes this really apparent the absence of Shadows is clear here so we probably don't need to point out the individual examples the car undersides make it particularly obvious it's also clear in the trees and in the background where toggling Shadows off removes any depth or self- shadowing from the trees leaves and the canopies if you're desperate for performance you should turn this option off but unlike other settings where there's sometimes no visual impact for a high performance cost in some games this one does have a visual impact and it is tied to Performance impact now we're comparing the presets we're going to split our Benchmark course into a few scenes and cover them in an order we discovered some really interesting things during this process for example the textures seem broken or extremely buggy where as the quality increases some of the textures get a lot worse additionally we couldn't find great reasons to play the game above medium settings which we'll talk about and show you so keep in mind that changing the presets changes every setting simultaneously so in this section we're not definitively looking at any one individual piece like we did with shadows but rather just the presets and note as well that resolution scaling once again was intentionally disabled for these tests although normally it would be enabled as part of the presets we made this table to show the different settings that are toggled when in different presets and you can see this in the game menu as well under the simple tab starting with a wide shot of the city the first visible difference is the lack of fog at the very low preset fog quality is simply marked enabled at every other preset and it doesn't change appearance between them but at very low the full screen Haze is completely gone frustratingly the fog gets in the way of most of the other settings we want to compare the next most obvious change is Shadows again with very low automatically setting Shadow quality to disabled we won't waste time rehashing that no Shadows means no Shadows at this distance though even the differences between quality levels with shadows enabled is unclear neither the draw distance nor the softness of the Shadows changes visibly so we'll check back in on that once we zoom in interestingly there's a clear increase in quantity of distant Tre trees when moving from the very low to the low presets but no change above that out of the given settings we'd suspect that trees are governed by LOD or maybe terrain quality but each of these settings is supposed to scale up with presets above low the highlighted power lines between the pylons crossing the river pop in sooner at medium and high but we haven't played much City skylines and it's unclear to us whether these highlights are even intentional either way they never show up in this recording with the reduced at very low although they would if we moved closer with a ton of qualifiers getting to the next comparison in this relatively flat City on a sunny day without any water related disasters at this distance at this resolution we haven't seen a meaningful difference between low and high presets so far there is one but the keyword here is Meaningful so let's keep looking continuing to pan around the city there are clearly visible reflections of bridges roads and boats on the surface of the river at the medium and high presets but not below that Reflections move from disabled with the very low preset to low at the low preset but evidently the lowest tier of reflection quality doesn't include these elements or at least not at this distance there are definite changes related to anti-aliasing between presets though moving from zero AA to FX AA by bumping the preset from very low to low dims some of the sparkling along the edges of Roads and Bridge wires but the fog is also a factor zooming in a single Street we can begin to see more subtle differences between the presets we'll freeze here for a second to point out some specific changes the shadow under this tree next to the sidewalk is slightly sharper at the medium and high presets than it is at low and the dithered edge of the shadow of the brick building at the right becomes sharper as well specifically at high the small yellow cars parked along the street cast Shadows down from their side mirrors but that's because the LOD setting erases their side mirrors at this distance also the handlebars on the red motorcycle are completely erased likely by AA at the two highest presets zooming along the street motion blur is non-existent with the very low preset but increases in intensity from low to high in addition to the intentional scaling of blur quality lower FPS at higher presets is a potential Factor here the developer specifically warns that quote motion blur can be very visually disruptive if you have low FPS end quote and that's just always been true moving to the very end of our course scaling and reflection quality is visible in the cars lining the street at medium and high the cars lining the street clearly reflects their own side mirrors while progressively more detail is added to the reflections on the side of this red car bizarrely the quality of the textures on the signs at the End of the Street get progressively worse with each step up and preset quality at very low the green and yellow billboard is crisp and readable but at high it's no longer recognizable as a billboard at all it just looks like a blurry mess other nearby textures suffer the same effect like benches and leaves on the trees we also saw this in the street textures when we were up close the back of cars and other objects this seems either completely broken as in they flipped the settings by accident or it seems to be an issue with rendering due to load intensity Al where this was captured on an RTX 490 so vram is definitely not a limiting factor and the performance got worse with the higher settings so we know it's not a universal thing it might just be this specific setting that's governing the texture quality in these scenarios overall isolating our analysis to this specific benchmark cor is this specific City we don't see any compelling reason to pick a preset above medium other than some Reflections even low doesn't look much worse than high in this area again we disabled resolution scaling for these comparisons so we're not taking that into account if you can you probably should disable resolution scaling because most of the time it looks completely awful it's using an FSR 1.0 implementation so you get a lot of flickering and sort of polka dot patterns as they pop in and out first off for conclusions here we need to address something that we saw that was very wrong in some of the posts online so I saw a a couple of Reddit posts and one steam review that claimed things such as toggling depth of field off boosting someone's performance from 22 fps to 122 FPS saw a steam review that claimed multiples of uplift from toggling depth of field off that's not true that's not how that works if someone got that kind of performance difference it's from something else and one of the things we noticed in testing is that the game settings especially if you do things like minimize or you have different uh selections for windowed versus full screen things like that it will change the resolution sometimes in ways that you don't expect it seems like a bug or in the very least it's just a bad way for user interface to behave but we did catch multiple times the game would adjust the resolution and more likely these types of performance increases people are talking about might be from going from say 4K to 1080 they just didn't notice that that changed or something else it absolutely was not depth of field that gained you 100 FPS it there's something else in there so just want to set the expectation uh that setting does help yes it doesn't help that much we're talking like less than 10% in most cases as for the performance impact looking at those charts again there's a huge Swain from very low to low so the developers have bad default here they should have more even space between these maybe they should add another named setting between them as a preset to give players a better balance without making the game look like trash so there's no reason they can't have something where they maybe Shuffle the names ultra low they add a new very low that's in between the current very low and low so that again people can get something that looks a little more presentable but doesn't uh have the same performance hit as just using Straight Up low settings there's a lot of room for tuning here in general uh you the developers in front of this they know it's a problem so they know they've posted they've said as much effectively as it being unoptimized and needing some work uh and the good news is that first patch did have anywhere from about 5% to let's call it about 10 to 12% on average for uplift with the exception being cards that were extremely low down the stack were we had one at 15 and one at 18% uplift now when your standard deviation run to run for something that 11 to 13 FPS depending on which patch it is is 0.4 that's that becomes a significant percentage of the total number so uh it's maybe not quite as as big as 18% but either way the patch helps more the worse the card is for performance the more sub30 it is so seeing these improvements it's a good first step it gives us some hope that there's things they can do to optimize and we think they can get there especially by tweaking some of the LOD settings uh the shadow settings the textures there's some room to play around and then volumetrics so lots of places they can improve if the base game is good that's the most important part and now it's it's a matter of getting it so that more people can play it without issues but hopefully this gives you a starting point if you were trying to get it to where you have a good frame rate but it looks a little bit better than just simply putting it on very low and moving on with your day that's it for this one thank you for watching as always to support us directly for this type of content because we had to delay our whole pipeline to add another day of testing just for the patch and we spent several days on it already go to store. Gamers access.net where you can grab one of our mod mats coaster packs shirts soldering and project mats or more and thanks for watching subscribe more we'll see you all next time

Channel: Gamers Nexus

Views: 392,361

Keywords: gamersnexus, gamers nexus, computer hardware, cities skylines 2, cities skylines 2 benchmarks, cities skylines 2 review, cities skylines 2 gpu benchmarks, best gpus for cities skylines 2, cities skylines 2 graphics optimization, cities skylines 2 settings

Id: l4DX6mUY78s

Length: 29min 0sec (1740 seconds)

Published: Sun Oct 29 2023