Gadgetory

so it's finally time to open up that cash of cash and see how it performs in games and welcome back to taking our city ladies and gentlemen this is brian coming back to you guys today finally with a video on level three cache and does it affect FPS not only that does it affect frame rates and frame latency that's the actual experience you getting on your monitor so for these tests I decided to test the G 32:58 at 3.8 gigs I then pitted it against the Xeon 12:31 v3 and I decided to disable hyper threading on this CPU as well as to cause so what we had here was pretty much an apples to apples comparison except the difference the main difference being that this one had an extra five megabytes of level 3 cache to use and I also decided to throw in the 5820k which has 15 megabytes at level 3 cache they keep in mind that this CPU is on a different platform so I kind of the main results that I was focusing on was these two which is pretty much an apples to apples comparison alright without further ado let's move on now to the benchmarks where I tested three games ARMA 3 Skyrim and battlefield 4 and they'll talk about a little bit more in depth what's going on in the benchmarks and then we'll move on to a conclusion so let's get on to the benchmarks now and I will put the specs of the rigs used in the description below so if you want to check that out then just pop down in the description and you'll be able to see that there though with all my benchmarks I'll be taking you guys through piece by piece and letting you know what's happening in every single benchmark and explaining things as best as I can because I'm sure there's a lot of people out there who are going to be like why what's going on here so I'll try and explain things as best as I can but that being said let's move on now to the first benchmark and that is battlefield 4 and now this was done with pre-rendered frames set to 1 also no vertical sync ultra on the test range at 1080p and so what we see here is that the really all these three CPUs impressed the results were really good on all three of these CPUs on this test I couldn't notice a difference when I analyzed it with the camera footage as well I couldn't really tell a difference I mean there was all all three of these CPUs did really well on this benchmark with the exception of the 5820k scoring a higher average FPS though we'll talk about that later and for what it's worth with these benchmarks I'm going to be focusing on the third 2:58 versus the Zeon mainly because they were done on the same motherboard same everything same RAM same exact same Ram speeds as a person the 5820k which is on a different platform the x99 and he just ddr4 memory though I still tried to get that as close as possible though and you probably wondering as well what is this this is a frame latency graph so what I've done here instead of doing 99th percentile or anything like that I've just taken every single frame and put it on a scatter plot and I've put the differences between the mean so basically any frame or any jerky experience will spike up out of proportion and you'll probably be able to notice that when it comes to gaming but honestly all all these all these frames here came under 30 milliseconds which is actually a pretty good score for what it's worth so battlefield 4 really there wasn't much of a difference here let's move on now to ARMA 3 we've got this again results that impressed on all three CPUs with the exception of the 32 58 as you can see here it actually the graph ended early for the 32 58 because that scored a lower average fps and we'll get onto that later with average FPS but what it's worth here all three CPUs didn't differ much from the mean in other words they all produced a pretty solid gaming experience though when I analyzed it with the camera footage I think that 32:58 exhibited a torn frame so that was something to come out of this graph ARMA 3 I guess subjectively and objectively performed better on the Xeon and the 5820k than it did on these 32:58 so I guess for ARMA 3 level 3 cache did make a little bit of a difference there considering it's the only factor and between the three CPUs in this - in these tests sorry let's move on now to Skyrim where we've got the 32 58 and the other 12 31 b3 in 5820k really putting out again a pretty impressive display here I mean we've got here the 1231 v3 scoring the worst here with a frame that differed from the mean by 24 milliseconds which again if you were playing Skyrim on any of these three CPUs was a 660ti you're going to get a pretty damn good experience we saw here the first part of this the benchmark was just absolutely butter smooth like that's almost like vsync even better like that was incredible so when I objectively measured this with the camera it showed the same three same thing so if you sat me down in the blind ABC test I couldn't tell you the difference at all between these three CPUs on Skyrim though another interesting test that I decided to throw into the mix was what if we turned on adaptive vsync would there be a difference as well because I'm sure there's some people who play RPGs with vsync on and so again we saw here that there really wasn't much of a difference again even so it probably produced a better result than it did on the previous benchmark so that was actually a really good eye opener there so essentially you know and two of these games are very CPU intensive Skyrim and ARMA 3 their CPU intensive for what it's worth and we weren't seeing that much of a difference I mean I'm a three did show three I'd say like 7% in average frames we'll get onto that later though let's look at the average FPS on battlefield 4 so what we had here was eighty nine point two versus eighty nine point three five so again the Xeon versus the Pentium in battlefield 4 didn't make much of a difference even the Pentium 32:58 edged out on minimum FPS so that's probably just variance though what we saw here was the 5820k pulling ahead in the higher average FPS and this was consistent when I did the benchmarks cuz I pretty much you know I double check all my benchmarks we saw that the 5820k score at a higher average I mean higher maximum which helped it pull a higher average in my opinion so that was pretty interesting so maybe in battlefield 4 a lot of level 3 cache does help give you a better gaming experience so that's something to keep in mind maybe I mean when I again when I looked at the camera footage I couldn't really tell a difference at all in this game looking at ARMA 3 this was the biggest result so this is probably single handedly this is the most important graph to look at in all these benchmarks is ARMA 3 because what we saw here was the 1231 v3 scoring Pro yeah like get off the top my head 7% difference in frames over the 32:58 the 1231 d3 so for what it's worth ARMA 3 made a difference and the funny thing about this is and this is where sort of opens up another question for me is is there a sweet spot for level 3 cache on games because I know when I tested the for 670 K that I had at four point six gigs as opposed to 5820k at four point two gigs the difference was pretty much like correlated in clock speeds so maybe there's a difference maybe there's like a sweet spot for level three cache that's interesting and in that being said it most likely depends on the game as well so if you're playing arma 3 like if you're just an ARMA 3 diehard fan you'll probably want to get something like a 4 7 9 TK that probably the best CPU bang for buck for this game or a 4690k that being said that I mean honestly though the 32:58 just clock at higher and you'll get an absolutely fine experience as well so anyway let's move on now to the next graph here this is FPS on Skyrim really this was the most you know one of the most insignificant results here like just yet I mean I didn't even put the adaptive vsync results in because they're pretty much identical and so we got here too I mean these are almost identical as well you get two extra FPS coming out a level 3 cache so only one game really made a difference in FBS and the frame latency and when I measured it with the camera didn't make too much of a difference though another interesting benchmark that I will throw into the mix is Cinebench when I ran Cinebench the actual center bench test itself didn't show a difference but the OpenGL tests showed a huge difference this was with the only the like again I had 2 cores 2 threads the the only difference was the level 3 cache and it made a huge difference on that OpenGL test so that was something interesting that came out of the benchmarks on Cinebench anyway guys with that being said let's move on now to a conclusion for you guys and we'll talk about pretty much what just happened so in conclusion the results were pretty anticlimactic I came into these bench expecting more of a difference from level 3 cache but the reality is there just wasn't that much of a difference between these 3 CPUs coupled with a 660ti now the biggest difference to be gained was from ARMA 3 where it showed like an 8% increase in frames which was pretty big and that probably raises another question of is there a sweet spot for level 3 cache with certain games and the main answer is it depends on the game that's the biggest thing that I got out of this every games coded differently and level 3 cache can make a difference though it really pales in comparison to other things like your core speed your core clock speed the IPC your core and thread count and the price of the CPU so level 3 cache ultimately isn't something that you should really take into heart when buying a CPU though it is pretty important for other things just for gaming it doesn't make that much of a difference now another thing is as to as I coupled this CPU with a 660ti and this is something that I'm going to stress from this video even though it's not really related is that when you're buying a low end CPU like the G 32:58 make sure you couple it with like a low end or a mid-range graphics card do not go coupling this thing with it like an online 290 X or GTX 980 like I've got a heap of comments now where people are telling me they're getting a crap gaming experience and they've got a 32 58 and they're coupling with a high end CPU like don't do that I've tested it in the past I've tested my pen in with the 780 lightning in the past and it was just a crap experience like I seriously had more fun playing games with the G 32:58 on a 760 than I did on a 780 lightning so it's something to keep in mind when you're going with a low in CPU just remember that you have to balance things out now dx12 might change that I don't know there's not many mainstream games or if any that used the X 12 at the moment so for what it's worth level 3 cache doesn't make too much of a difference for games and if you have any comments or questions drop a comment in the comment section below and I'll get back to you as soon as I can and if you haven't already then hit that like button as this video took me a long time to do in terms of compiling and double-checking the results and I'll catch you guys in another tech video very soon and you can hit the subscribe button if you want to check me out on social media and I'll catch you in another tech video peace out for now bye well conditions rubbing your feet on carpet whatever producing us you know just a little bit of static electricity is that enough to damage computer components