in my last video we were looking at if the 4070 TI super was worth $200 more than a 4070 super and price to Performance wise right now today it isn't as from a value proposition except for the fact that that jumps you up to 16 GB of vram over the 12 GB of vram but the 470 super was certainly holding a better price to Performance metric but again that vram situation but if you do decide to spend the additional money to get the 16 GB of vram maybe you hang on to your gpus for two or three generations you're like 12's not going to cut it cut it for that long well then you now would want to be like considering okay I'm spending 800 on the 470 TI super should I spend another $200 extra on the 4080 super does that make any sense and that's the comparison we're focusing on in today's video the 16 GB 480 super against the 16 GB 470 TI super notice this time they're both 16 GB products so you're not getting additional vram if you spend the extra $200 and is it even $200 extra well on msrps it should be uh the 4080 super should be $999.99 and I haven't seen one in stock in that price for a long time although one just popped up available a few minutes before I filmed this video I have a feeling if I film this uh a few hours later this one will probably be gone again uh cuz the MSRP models have been few and far between and most of them have been sitting in stock closer to the $1,200 price point that 480s were sitting at all all along and remember the 4080 and the 480 super perform the same it's within a one 1 to 3% variance so they're basically the same product so buy whichever one's cheaper but anyway um so in other words if we're looking at the value calculation with this we could look at it as msrps but also what what they actually cost on the current market So currently you're looking at spending 25% more money if you want to buy the 4080 super over the 470 TI super uh sorry that would be $200 more which is a 25% % price increase um now the problem also is that like I said most of the models are closer to $1,200 which would be a 50% price increase and to to be clear the ti the 4070 TI supers are very much widely available at their $800 reference MSRP several models in stock right now and a few more that are just a little bit higher if you want to consider some other cooler options so that is definitely the case so in other words right now you might be spending 50% more money if you go with that 4080 super and by the way let me talk about this for a second in my last video talking about the $800 versus $600 on the 470 super and 470 TI super I actually had numerous comments telling me I'm doing my percent increase wrong uh on these calculations so let's talk about that for a second because you can measure percent differences two ways uh depending on which one you set as the Baseline that you're considering to be 100% And what's important isn't which way you do it it's whether it's in line with your performance calculations both need to be done the same way so if we look at this screenshot here you can see that my 480 super is uh showing up as 118% what does that mean uh the 4070 TI super setting as 100% I'm I'm using these numbers as my calculations and so what I'm doing is I'm calling 40 FPS the 100% Baseline and I'm measuring the 47 FPS as the performance relative to that which would be 118% which is an 18% increase over the Baseline value of 40 so that doesn't really matter so much as far far as which way you do it because I could have also set the 480 super as 100% but then I could have set the 470 TI super as a percentage of that and either way works what matters is when you do your price calculations and your performance calculations you do it consistently you might be like does does it really matter it does because notice that 1 1200 compared to 800 is a 50% % increase over 800 because it's a $400 difference and $400 is 50% half of 800 so you increase 800 by 50% of 800 to get to 1,00 if you did it the other way around 800 out of 1,00 you now get 67% in in other words 2/3 uh because 800 is 2/3 of 1,200 the the the uh $400 price difference if you consider 1,200 to be your 100% Baseline uh well 400 is only 13 of 1200 or 33% is of 1200 so it's a 33% difference if you look at it this way but it's a 50% difference if you look at it this way and it doesn't actually matter which way you do this as long as you do the performance and the price the same way sorry for my math rant I'm actually a math teacher so I got a little bit frustrated when people were calling me out on my math calculations there in the comment section you can do it either way but you wouldn't uh you wouldn't say it with the same words you would say that the 480 super at $1,200 would be 50% more money than the 4070 TI super because you're comparing it to the 40 470 TI super if you were looking at it this way you would say the 4070 TI super costs 67% of the price of the $1,200 480 super see how you phrase that slightly differently anyway uh I think I got way sidetracked there sorry for the math lesson let's run the benchmarks and I'll give you some final thoughts at the end we'll kick off testing an avatar friend tiers of Pandora and at 1080p ultra settings both gpus are doing great 107 FPS versus 123 this is one of the best looking new titles out there it does use Ray tracing at all presets there's no way to fully turn it off and just a friendly reminder about the important note down at the bottom of the screen mentioning that when you capture at 60 frames per second it can introduce uh motion jutter and things like that to the recording so please judge smoothness by the frame time graphs and the 1% lows anyway let's move on up to 1440p ultra I think more people would be targeting 1440p with these gpus and here we're seeing the ti super at 72 FPS and the 480 super at 87 so both pretty good and is a 21% lead for the 480 super 16% lead and the 1% lows at 72 versus 62 and especially the ti super could maybe do for a bit of a higher frame rate uh in a first-person shooter Style game I know for me even in single player games you know that 80 to 100 FPS range feels a little nicer in the firstperson shooter style so uh if we kick on dlss quality that gets the job done the 470 up to 108 and the 4080 up to 128 giving it a 19% advantage and the low is at 99 versus 86 a 15% Advantage but if you don't want to use upscaling at 1440p although it does look pretty good you can just turn down to high settings Nvidia hates this one simple trick of stretching out the life of your GPU don't turn on ultra settings anyway uh we go to 114 FPS versus 99 so good high high refresh rate experience with no upscaling a 15% Advantage for the 4080 in both the averages and the 1% lows which are 93 versus 81 but what if you're trying to buy these for a 4K resolution I mean we're talking $800 to $1,000 MSRP products well 4K Ultra is going to be a bit much it's certainly playable but probably not the way You' choose to play if that makes sense uh 47 FPS versus 40 gives the 4080 and 18% advantage in the averages and a 177% advantage in the 1% lows at 42 versus 36 so what if we kick on dlss quality which at 4K resolution I mean it's not always exactly just free performance from an image quality perspective but it's pretty close looks quite good at 4K you got a lot of pixels to work with and now our frame rates jump up considerably to 77 and 65 respectively giving the 4080 an 18% advantage in the averages and 16% advantage in the 1% lows at 66 versus 57 both gpus very playable there although again if you want to stay at Native resolution you could try reducing settings however at high settings it I mean we're a little above or below 60 now depending on which GPU you're using at 63 versus 55 15% advantage to the 4080 uh but I don't know if it's quite the high refresh rate experience you'd prefer in a first-person shooter of course if you're sitting back with a controller on your couch this is probably fine on a 4K TV but uh with a mouse and a firstperson shooter kick on some dlss quality and if you're using high as your Baseline of settings we are now able to get a high refresh rate experience that I would say uh would be uh probably how I'd be playing it on these gpus we're at 100 FPS versus 86 which is a 16% Advantage for the 4080 in the averages a 177% advantage in the 1% lows at 83 versus 71 but let's jump into one of the other best looking new games pushing graphics and while this game does have a crazy path tracing mode we'll take a look at that in a minute first we'll start out at the high preset there is no Ultra preset in this game this is the highest non-r race preset and now we see the 4080 only winning by 10% at 81 versus 74 at 1440p uh but you know that's good you know frame rates for this game in any case it's it's a pretty slow paced game and uh that's a pretty high refresh rate experience however if if you're looking for even higher refresh rates you could kick on dlss quality and uh interestingly this game runs dlss at all times although you can run it native daa resolution um which is kind of interesting there's no there's no TAA pathway but that's a bit of a side note here we're seeing a 12% Advantage for the 480 at 118 versus 105 uh both of them offering a nice high refresh rate experience here with dlss quality enabled but if you're already at high refresh rate experience this is a game that offers frame generation and if you kick on dlss 3 frame generation when you already have a high base frame rate then the latency hit isn't too bad you can see down at the bottom of the screen uh and the uh overall image quality is still pretty good because there's not large gaps between frames when you're interpolating them so you get a very high uh motion fluidity keep in mind that these frame rates would not be equivalent to achieving those frame rates without uh um frame generation interpolating frames if you were able to hit those natively speaking of native how about 4K native resolution at the highest settings well once again sort of like we saw in Avatar it's a bit much for these gpus it's playable but again probably not the way you'd want to play it we're at 48 versus 42 which is a 14% Advantage for the 480 and a 19% advantage in the 1% lows at 37 versus 31 so I think it would definitely make sense to kick on dlss quality and if we do that again a large performance bump occurs and the uh frame rates are now at 76 versus 68 respectively which is a 12% Advantage for the 4080 and a 177% advantage in the 1% lows at 56 versus 48 and so the game uh you know this slow paste of a game I think that this is perfectly fine but also you can kick on frame generation in this game again if you're at a high enough base frame rate and for me if you're over 60 it's probably fine um but everybody experiences it a little bit differently so of course try it out for yourself self King on frame generation gets the motion fluidity up to 102 frames per second and 89 frames per second respectively a 15% Advantage for the 480 but again keep in mind that the responsiveness and image quality is not quite as good as if you achieved those frame rates without frame generation but what about the path tracing mode I know you're probably not buying these gpus for 1080p but let's just throw it in here to see how they do at 1080p RT High settings that's kicking on all of the path tracing everything maxed out um we're a native resolution no dlss but it's only 1080p the 480 can hit 61 average here versus 54 on the 470 TI super which is a 13% Advantage so they're both able to hit around that 60 FPS Mark at 1080p what about at 1440p well without dlss you're not going to have a good time however if you kick on dlss quality which is what we're measuring here no frame generation yet uh we're seeing right around that 60fps experience on both gpus one a bit above one a bit of below giving the 4080 a 12% Advantage at 65 versus 58 FPS averages 48 versus 43 in the 1% lows but again when you're around that 60 FPS Mark frame generation can be a pretty good experience to smooth out uh the motion on a high refresh rate display again not giving you the full uh you know responsiveness of the high refresh rate experience but we're getting 107 versus 95 FPS of motion fluidity now which is a 13% advantage to the 4080 and um again that's with the path tracing stuff maxed out dlss quality pretty good experience there but can these gpus handle 4K path tracing and I mean if you go all the way down to dlss Performance mode it's 50 FPS versus 44 14% advantage to the 480 DLS performance mode does not look too bad at 4K resolution um so this is maybe an acceptable experience but we're certainly not hitting 60 FPS so um I don't know guys uh you could try kicking on frame generation but like I said for me I'd like to have a 60fps baseline before I kick it on if you look at the average PC latency down at the bottom of the screen those numbers are a bit higher than I'm personally comfortable with when using a mouse if using a controller it doesn't bother me as much I guess I either just naturally expect more latency or there's less of a on toone responsiveness but anyway you can try frame generation here um to smooth things out on a higher refresh rate display let's move on to another path tracing test with cyberpunk at 1080p RT overdrive mode neither GPU hit 60 FPS in the built-in Benchmark run I'm seeing 54 versus 49 certainly playable it's a 10% advantage to the 4080 177% lead in the 1% lows at 42 versus 36 but uh you know it's it's definitely not a high refresh rate experience but I think people are going to be more interested in the 4 1440p results so rather than kicking on dlss at 1080p let's just move on to 1440p and with no dlss uh we're at 35 versus 29 FPS averages um which is not great so 21% lead for the 480 here in the averages and the 1% lows but it's kind of a pck victory or uh you know it it's just not the way you'd want to play the game so if you kick on dlss quality which at 1440p can still look quite good uh we're now getting a much more acceptable experience at 66 versus 59 FPS 12% lead for the 4080 and I think both gpus are perfectly playable now with path tracing and dlss quality is not too much of an image quality compromise here especially with Ray reconstruction helping some of the temporal stability although in this game it can have a a little bit of a weird look on faces I've seen but anyway um when we are able to achieve that kind of 60fps is Baseline I think kicking on frame generation can make sense you can see the PC late laty uh does go up compared to our last test without frame generation but the motion fluidity would be a lot higher at 109 versus 97 FPS here we're getting a 12% lead for the 4080 and a 14% lead in the averages and I have a feeling a lot of people this would be the settings you'd use on a 1440p display on either of these gpus to play this game if we move up to 4K resolution again kind of like we saw in Avatar not Avatar Allen wake 2 uh dlss performance mode with path tracing gets us playable but not over 60 FPS 53 versus 47 133% advantage to the 480 and the averages 15% advantage in the 1% lows again performance mode upscaling can be acceptable here we're just not getting to a real high refresh rate experience um you can try kicking on frame generation here the P the P average PC latency in cyberpunk doesn't seem as high as it is in uh Allen wake 2 every game can be a little bit different there uh so this I don't know this might work out for some people and not others there's also image quality concerns the lower your base frame rate is because the interpolation is more difficult we're seeing 73 versus 60 of motion fluidity now but not responsiveness uh 22% uh lead in the averages there but let's jump into some Unreal Engine 5 testing Lords of the Fallen is the most demanding Unreal Engine 5 game that I've been testing out and so I'll even look at it at 1080p where at ultra settings we're seeing 95 versus 84 so good results um but again that's only 1080p uh 1% lows are 71 versus 60 so 18% Advantage there and 133% advantage in the averages uh for the 4080 what about 1440p resolution you can see how demanding this is 1440p ultra sees the 470 TI super only hitting 62 FPS average which sure it's playable but it's hardly High refresh rate experience on an $800 GPU but 71 FPS on the 4080 super giving it a 15% lead in the averages and a 22% lead and the 1% lows at 55 versus 45 but Unreal Engine 5 really does seem to be designed with upscaling in mind it performance scales massively with resolution so if you lower the internal resolution using dlss quality we now see frame rates jump all the way up to 93 and 103 respectively giving the 4080 and 11% advantage in the averages but an overall High refresh rate experience now that 1% lows are 73 versus 65 which is a 12% lead for the 4080 what if we move up to 4K and like we've seen uh this engine you know per the resolution has a huge demand on performance and both of these gpus often are falling short of 60 FPS at 4K native resolution at ultra settings even without you know Ray tracing and all that we're seeing 41 versus 35 which is 177% lead for the 4080 14% lead and 1% lows but again probably not the way you're going to play the game so if we kick on dlss quality again Unreal Engine 5 seems to scale massively with resol rendering resolution and performance jumps all the way to 61 and 71 respectively that is a huge performance jump for kicking on dlss quality here uh that's uh so now at 71 versus 61 FPS we're getting a 16% lead for the 4080 and a 177% lead and the 1% lows at 55 versus 47 I think we got a pretty good experience here there's no frame generation available in Lords of the Fallen currently but let's move on to RoboCop Rogue city which is another Unreal Engine 5 game and is less demanding at its Max settings here the max settings are called epic settings and uh 1080p uh you know we're kind of almost hitting CPU limits at 1440p we're still getting a pretty good experience here at 86 versus 79 9% Advantage for the 4080 uh so not a massive lead here for the 4080 but pretty good experience on both gpus um let's just jump all the way up up to 4K resolution where again you can see that the the demand by moving up to 4K is considerable and at Epic settings both gpus are once again falling below 60 FPS here we're at 50 versus 43 playable but probably not how you'd want to play a firstperson shooter 16% Advantage for the 480 in the averages and a 19% advantage in the 1% lows at 44 versus 37 but this being Unreal Engine 5 if you kick on dlss quality it again scales performance massive ly and we're seeing 77 versus 68 now uh for our average frame rates giving a 133% advantage to the 480 super and if we look at the 1% lows we're seeing 65 versus 57 giving a 14% advantage to the 4080 super and since we're over 60 FPS this game does offer frame generation and I thought I'd kick it on and so our motion fluidity now increases to 105 frames per second and 93 at frames per second 13% lead to the 480 but again keep in mind that the motion fluidity offered by frame generation is not the same thing as the increased responsiveness you would get from achieving this frame rate naturally but the PC latency here does feel pretty good to me at these settings and frame rates let's go ahead and move on to Starfield at 1440p ultra settings uh we are getting 88 versus 79 FPS that is an 11% advantage to the 40 80 super and a 99% lead in the 1% lows at 63 versus 58 I'm using the section of new Atlantis for the test which is more demanding than a lot of the kind of you know combat areas and things of the game but less demanding than a few others though so I think it's pretty fair test uh if we move up to 4K resolution though at ultra settings we're now seeing the 4080 able to hit 60 FPS the 4070 TI Super Falling a bit short at 53 a variable refresh rate display it probably wouldn't feel massively different though a that's a 133% Advantage for the 480 and a one uh the 1% lows are a 10% Advantage for the 480 at 45 versus 41 but again in a firstperson shooter Style game those probably weren't the frame rates we were looking for so what about dlss quality which looks quite good at 4K and now we're jumping to 80 FPS versus 72 respectively giving an 11% lead to the 4080 the 1% lows are at 59 50 giving an 18% advantage to the 4080 super and we overall have a pretty good experience here however this game does offer frame generation so if you want to take advantage of the motion fluidity of a high refresh rate display we had a high enough base frame rate that frame generation could make sense increasing the motion fluidity to 111 and 99 frames per second respectively 12% advantage to the 4080 1% lows are now 92 versus 84 which is a big jump and I think uh the 1% lows here can be very CPU bound and frame generation can push through that just keep in mind that again it's not quite equivalent to achieving those frame rates without frame generation what about balers Gate 3 well at 1440p we are CPU limited which is why the gpus are within margin of error of each other at 1% advantage in the 1% lows uh to the 4080 3% advantage in the uh averages to the ti super but again that's just variance uh you know in CPU limited situation uh a 1 to 3% variance is is pretty common so basically that's a tie in a CPU limited situation nothing to see here move on to 4K Ultra though and we are now more GPU bound and we're seeing the 4080 super able to pull ahead by 16% at 94 FPS versus 81 both of these gpus are putting up a ma very good frame rates for this style of game so basically yes you can Max this game out natively at 4k on both G gpus no no trouble uh let's go ahead and move on to a different game engine as well this is the re engine from Capcom with Resident Evil 4 remake and 1440p Max settings do include some Ray traced Reflections although they're not super high resolution this game does like to use a lot of vram and all that but frame rates are good here we're at 157 versus 136 15% advantage to the 480 super 11% lead and the 1% lows let's crank things up to 4K so at 4K settings we're still even without any upscaling able to hit good frame rates 96 versus 81 it's a 19% advantage to the 480 super and the 1% lows are 87 versus 73 again a 19% advantage to the 480 super you see that this uh this engine game certainly seems uh very performant at 4K even compared to a lot of the other games we've tested here so far now let's go ahead and test out one more game with Call Of Duty Modern Warfare 3 and I'll be testing the basic preset assuming that you'd be uh turning down visual settings to increase frame rates for competitive purposes and that pushes our frame rates at 1080p all the way up to 352 versus 328 which is only a 7% advantage to the 4080 but again the lower your resolution is the more likely uh you are to be slowed down by other things in the system or game engine things like that not fully GPU bound anyway uh if we move up to 1440p basic settings we're now seeing a 10% ad vage for the 480 super although a 20% advantage in the 1% lows the averages are 282 versus 256 respectively so certainly high enough that if you didn't get the kill it's it's a skill issue not the not the hardware guys and the 1% lows are 21 versus 168 uh so again a little bit more meaningful lead there in the 1% lows and if we go up to 4K basic settings we're now seeing a 15% Advantage for the 480 and 21% lead in the 1% low so so again it's growing as the resolution increases the average frame rates at 180 versus 157 are still quite High although you could definitely turn on something like uh upscaling here dlss quality to boost frame rates even more and still look quite good uh even dlss performance mode in competitive game Let's pop out for some final thoughts all right so final thoughts if we go ahead and just look at this relative performance number on the 4080 super and I just kind of quickly scroll through the results what you'll notice is that generally we're getting numbers between 10 and 20% 20% is pretty rare we're usually somewhere in between 15% seems to be you know a pretty uh reasonable number to call somewhere in the middle uh you know both in frequency and results so basically I think calling it roughly a 15ish per performance increase for the 480 super is pretty reasonable so then you could do some value Val calculations but I also think that value calculations only gets you part of the picture when you're considering two different gpus for example when I was comparing the 12 GB GPU versus the 16 GB GPU you have to think about okay there's longevity versus now and things like that but remember 480 super and the 470 TI super from a feature set standpoint and from a vram standpoint and everything like that else it there's really not much in it it's just a performance and price difference so uh with that in mind is this level of performance getting you a meaningfully different experience so you know in the hardest situations we have right now uh we're talking like path tracing modes and things like that does it get you a meaningfully different experience well you're getting just under 60 FPS on your 4070 TI super you're getting a little over 60 FPS on the 4080 super but it's not like uh you know going from a 60 to like a high refresh rate experience I would call these in basically the same experience from like if you turned the frame rate counters off and you were just playing the game would you notice a a huge difference jumping from one PC to the other if you had an AB blind test here and you might be able to tell the 4080 super was running a little bit faster but guys when it's like a 12% performance Gap that type of thing like it it's I don't I don't think we're getting a meaningfully different thing here now what about a 4K resolution where you can where you really need every frame you can get well the problem here is for the most anding 4K titles like I said path tracing modes even with performance level upscaling uh neither GPU is delivering 60fps and so if we get even more demanding path tracing games coming out over the next few years I'm not sure these are 4K path tracing gpus so in other words let me put it like this I don't think the 4080 super can really do anything meaningfully different than the 4070 TI super so uh if you look at that it from that perspective I don't think that when it's got the inflated price tag of $1,200 it makes any any sense at all you're paying like I said if it's at the you know $1,200 price point 50% more money to get 10 to 15% more performance occasionally maybe 20% but that's pretty rare now if it's available at its actual MSRP of $11,000 you're paying 25% more money to get maybe $15 more perc more performance so it's certainly worse value but the scaling isn't awful keep in mind that in general throughout the history of the GPU Market there's usually diminishing returns uh for a performance to per dollar standpoint as you go up the more expensive gpus in the product stack so in other words at its MSRP of $1,000 it's not a ridiculous thing to do to spend the additional money but it is definitely better value to get the 4070 TI super and then just I I just want to remind you that the 470 super non TI at $600 if you're just looking at a performance per dollar calculation offered better value than the 4070 TI super so you got the 480 super offering worse value than the 470 TI super which offers worse value than the 470 super so if you're hunting value stop at the $600 price point or you can also consider some AMD Alternatives depending on where your priorities lie on you know Ray tracing and upscaling and all of that but explore my my head-to-head comparisons with AMD gpus if you want to go that route uh if you want to move up to 4K path tracing situations I think only the 490 is going to offer you that and yeah it's really expensive right now um and you know maybe then just wait for what the high-end gpus and like the 50 Series have to offer a year from now that kind of thing but yeah so my overall thought is if if it is at if they're both at their MSRP the 4070 TI super is still better value but it's not ridiculously slow um but if you're thinking about spending $1,200 on a 480 super I really think the 4070 TI super is probably just makes a lot more sense anyway I hope you guys found this video useful and or interesting and I hope uh you know actually huge thank you to viewers commenters even if you're criticizing my math we can have a math conversation subscribers and especially channel members who have directly funded the channel on a monthly basis by clicking the join button that is absolutely huge thank you it helps to get the motivation to put GA these massive head-to-head comparisons which do take quite a bit of time to put together running the benchmarks and editing and all that so huge thank you but I totally understand that not everybody's in a financial position to do that so that is totally fine and I hope all of you have an excellent day
