Gadgetory


All Cool Mind-blowing Gadgets You Love in One Place

How much does AMD REALLY bottleneck?

2016-10-06
this shirts from last year and it's way too big on me but I really like it because it tells you exactly what you should keep in mind when you go into the comments of today's video so when I did the five hundred and fifty dollar gaming PC and I chose AMD as my processor specifically the 8320 everyone kind of flipped their and I knew as I was building that system I was like I'm gonna follow this up with a bottlenecking video it's been a long time since I've done one admittedly all the way back to the 3770k that was still new when I did a bottlenecking video if that gives you any idea of how long it's been so we're gonna do today is we're gonna take some high-end Hardware play some games and we're going to take the 8320 an we're going to play some games both of them on gtx 1080s we're gonna see how much of an FPS difference there really is today's video is brought to you by me I made it I brought it to you all I'm askin is that if you're new around here if you've been lurking for a while why'd you go and hit that subscribe button we'd like to join you here to the cool kids club showing that you have some of the best taste in tech youtubers and all of the internet and Who am I kidding this channels mediocre at best but I digress either way it's still like you to hit that subscribe button and hang around for a little bit but with that said let's go in jump into today's video so before we turn and get into some games there is some methodology we have to get out of the way here our baseline is going to be done on an Intel i7 59 30k overclocked to 4.4 gigahertz it's got 20 800 megahertz ddr4 HyperX predator RAM in there it's running on a x99 for the win and of course it is water-cooled now some of you are already gonna be like that's not fair because that's not representing what the average gamer is going to have I'm not trying to replicate what the average gamers going to have I'm trying to remove any possibility of any choke points or bottlenecks in the system to see exactly what the GTX 1080 is capable of when the CPU and all that is not gonna be holding it back in any way now the point of today's video is not to say that if you have an AMD system that you won't be bottlenecking or slowing down your FPS the question is how much so that's the point of today's video if you've gotten any other impression out of this video then I suggest you watch it again because that is not the goal here transition we're going to start off here with a little bit of 3d marked fire strike normal because normal tends to have a higher score which lets the GPU ramp up go as fast as it can because it can pump out a lot of frames in 1080 series we're going to because this is a base test here we're just going to let the test go we're going to see what the GPU score is and the combined score to kind of compare what the differences are and then we'll move on some games all right so one the 59 30 K we got a twenty two thousand 194 graphics score that's up slightly from the score I got when I did my initial testing it was like twenty two thousand ninety seven or something like that I mean we got a combined score of eighteen thousand six seventy so that's our baseline alright so Metro last light here super sampling is on tessellation all that Sun normal and then the settings are on very high so we're gonna let this test run and check the FPS so our average framerate on that was ninety three point seven six with a minimum we're gonna look at minimums here where they're available with a thirteen point nine to Metro last light though the minimums are kind of misleading if you will because when it starts it goes in single digits sometimes okay so next gonna be doing here some rise of the Tomb Raider we're doing this in dx11 mode not 12 because AMD system is running on Windows 8.1 so I can't test that in DirectX 12 maybe that'd be a future video how much does that alleviate some bottleneck but I think we've already talked about that a bit anyway here are the settings SMA and all that junk so let's go ahead and run the benchmark and then we'll see what again what our average FPS is rise of the Tomb Raider overall score 130 0.06 so 130 FPS was our average so here are the settings right here ambient occlusion hi everything's high and very high no reflection MSA population densities all up and then be sync off TX a is off of NSA's at 2 X and FX AAA is on of course we're at 1080 so those are the settings we're going to apply and we are going to do the beginning mission with Franklin where he has to repo the car I like that because it's on a track and it's the same every time I do it you can see the FPS counter up there in the corner obviously we're getting a lot of FPS here so we're going to do once we start the driving I start my fraps benchmark and then it will measure the FPS while I'm playing okay so we got rip fraps running up here I'm going to let it do its benchmark while we go and then at the end of this when we get to the parking garage we'll stop it and we'll see what the average FPS was all right average of 156 FPS is what that red number was alright next up is shadow of Mordor preset is on ultra of course we're running 1080p with unlimited FPS and we're going to do the benchmark and then were going to follow the benchmark with actual gameplay to see what kind of variants there is there too because I know I'm not sure if this benchmark is as rough on the graphics and CPU as playing the actual game would be so let's run this see what the settings give us and then we'll play some games team all right average FPS 160 5.21 max of 233 and a minimum of 98 technically 99 it was 98 point nine four all right so it's going to do a game and see if we get those same type of FPS I don't think we will I believe in the game I think it's limited but I've spent a long time since I played this game I'm not sure yeah so just like I thought inside the game here we are limited to 100 FPS okay so as expected amp es has just locked at 100 so I guess we're going to see if we get any less than 100 on the AMD system okay so as we start the Metro last light test here I put some stuff up on the screen for you guys to check out we've got the GPU temperature GPU usage and GPU frequency as well as the CPU temperature and the CPU usage on the bottom so what you kind of want to look for now going forward on this test with the AMD system is as the CPU usage goes up you'll probably see the GPU usage start to go down because the GPU is having to wait for the CPU to can finish whatever tasks it has going in its queue but so far though it doesn't look like the GPU is bouncing around too much I mean it's sitting in the upper 90s it went down to the 70s for a second and back up to the 90s and you can see the temperature is going up on the GPU and the frequency is staying pretty steady so we'll just let this run and then we'll compare it to the first result on the i7 system so here's a score eighty five point two one a minimum frame rate of one point one seven there were some pretty there was like two or three major stutters in there during explosions and that's obviously where the minimum frame rate came from but yeah so here's how those compared to the i7i sunrise the tomb raider you can see the GPU usage in percentage right there is down quite a bit compared to the CPU usage going up so you can see the CPU usage is sitting right around seventy-five to fifty is bouncing around quite a bit but as the CPU usage goes up you'll notice the GPU usage goes down so you can see our here our overall score is a ninety two point six two and that's how this compared to the i7 system here so you can see there's a difference question is how much I don't know I'm actually putting out the differences here I forgot what the scores were so I'm going to kind of learn along with you all right same test here with Grand Theft Auto 5 ok the average on this one was 91 so that came down a bit obviously from the i7 system which is funny because the CPU wasn't pegged but the GPU wasn't going to 100% either alright shout out more doors up we're gonna run the benchmark and then just like before we're gonna run the game okay so here's before and after a shadow of mordor average FPS 120 5.9 max FPS 205 point 6 1 and minimum FPS 81 and that's how that compared to the i7 system now in the game I'm pretty sure we're going to see the same results as before just a cap of 100 FPS even with the AMD system if we are still sitting at that hundred FPS number CPU usage is down in the 40s and high 30s GPU usage is in the 60s ok one last test to do that is the fire strike score and based on what we've seen so far across the board we would expect to see the GPU score on fire strike fall how much I don't know we know the combined score is definitely going to be lower because of the fact that I seven system is a extreme you know x99 motherboard with a 12 thread CPU in there that is overclocked so I obviously the combined score will be lower here but we care about is the GPU score is the CPU causing GPU to slow down I mean it is but the question is how much okay so here's the final result eleven thousand seven twenty nine on the combined we expected that to come down it came down about seven thousand points but the graphic score only came down a little under four hundred points there as you can see twenty one thousand eight thirty three is what we got versus of twenty two thousand two hundred something on the i7 so it really wasn't that much different now one could almost deduce from the firestrike score is that the 8320 is not really bottlenecking the gtx 1080 now the reason why that test shows very close scores compared to the actual games is the fact that fire strike is optimized to stress the GPU and put all of the tasks on the GPU with the exception of the combined physics with the CPU and the CPU only test that's why the combined score or the total score is much lower now in the case of games what's actually happening here is the CPU is taking longer to do its cpu specific tasks like figuring out where characters are where to put these things on the screen what's actually happening in the game because it's not on a track and preloaded like synthetic benchmarks so the GPU spent some time waiting which is why it doesn't spend all of its time up in the ninety ninety five ninety nine percent usage range because the CPU is handling its overall job a lot slower again as I said in the beginning though today's video is not about whether or not AMD does or doesn't bottleneck I think it goes without saying that it does create a choke point in your system if you match an AMD like 8320 or 8350 with a very high-end modern GPU like the GTX 1080 or god forbid even like a Titan Pascal now the percentage of what your games are going to slow down is going to depend on the optimization of the game itself and how much CPU it really relies on there's obviously very CPU dependent games out there like City skylines I think it's called but forgive me if I got that wrong obviously those games like Daisy HN h1n1 whatever those types of games use a lot of CPU because there's a lot of things it has to calculate that's happening in the world because the world is handled by the CPU the eye candy is handled by the GPU so that's why things like fire strike didn't show nearly as much of a slowdown obviously I don't think you should buy a 1080 and stick it in an AMD system because you're going to get a pretty diminished return on that investment but I was just kind of curious after I did that last video how much does it really slow it down you put in a mid-range GPU like a 950 or a 1050 volt if it doesn't exist yet or a 10 60 or something like that you're not going to see as much of a bottleneck but at least as you did with a 1080 but the chances are they're being a few percentage points the difference in fps is definitely going to be there because again the CPU is still having to handle those same tasks that I was having a hard time doing with the 1080 anyway what's happening is you put a lower-end GPU that can't render those massive amount of frames anyway then it's not waiting for the CPU because they're operating on a more equal level and in tandem all right guys time to go hope this videos helped as I said careful careful if you go down into the comments section of this video it's going to be a flame fest I don't even think I'm going to visit that myself I did the video I did the way I wanted to do it and the information is there I take you for what you will time to go guys as always thanks for watching and I'll see you in the next video
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.