Gadgetory

welcome back to harbor unboxed today's video it's been ever do over if I'm honest if you missed it roughly two weeks ago now I decided to bust out my own house wall core i7 test system install a GTX 1080 TI on it as you do and see how it compares to the new eighth generation Core i7 80s arcade in today's games player benchmarking did go into that one but you guys seem to really enjoy the comparison the only criticism though being that while I did overclock the older 4770k I didn't overclock the new 8700 K since we were already GPU bandwidth the GTX 1080 and a few examples that I gave I knew that the most part overclocking a d7 okay wouldn't really change much still quite a few passionate viewers dependent that I'd go back and retest with the 8700 care of a clock so that's what we've done I've also added some 1440p results as well and this is something that many of you also requested for this test we have a number of CPU configurations tested at any peer using sort of medium mid-range quality settings to ultra which is maximum the ultra preset as well as 1440p again using the ultra maximum quality preset the 4770k numbers have been updated with better ddr3 memory it's still 2400 stuff but the timings are a bit tighter and in some of the tests this has really helped I'm now testing with G skills sniper series 16 gigabyte ddr3 2400 CL 11 memory and I've also increased the overclock frequency by Turner megahertz to 4.8 gig arts for all the tests for a direct comparison I've also clocked the a7 ok at 4.8 gigahertz my chip does have no issue at 5 gigahertz but as I said I really wanted to do a clock for clock comparison here so I do clock it at 4.8 gigahertz and this is probably a more realistic overclock anyway it seems like most people with an 87 ROK are achieving around 4.8 gigahertz the big reason for this update has been to pave the way for my core i7 2600 so that should be interesting for now though let's check out the updated results like all good benchmark sessions we're going to be kicking things off with ashes of the benchmark very few gamers may actually play this game but that doesn't change the fact that from a technical standpoint it's a marvel and it puts most other games to shame it's also a perfect example for showing the relationship between the CPU and GPU and how quality settings impact to that the high quality preset which can be considered more of a medium type setting relaxes the demand on the GPU and therefore the CPU can stretch its legs a bit more even so overclocking a d7 arcade doesn't provide that much more in the way of extra performance so despite using the higher quality preset we're still able to find the limits of the GTX 10 atti clocked at 4.8 gigahertz the 4770k was good for a hundred and seven FPS on average and that meant overclocked 37 ok was 30% faster and that's a seriously big margin right there however jumping to the crazy preset which applies the maximum in game visual terrain to a serious GPU bottleneck for most of the test configurations the only exceptions here include the 4 threaded 7600 K and the lower clocked stock 4770k comparing the overclocked configurations the aether market is now just 4% faster for the average framerate moving a 1440p though sees the playing field completely neutralized and now there is no distinguishing between the overclocked 4770k and the 8700 k and remember we are still using the gtx 1080 TI for this test so in ashes of the singularity the 87 era K was up to 30% faster but under realistic conditions that see game is pretty much GPU bound there's likely to be little to no difference in this title next up we have Assassin's Creed origins and here are the 1080p medium quality preset results not much in it really after saying quite shockingly overclocked in the 87 arcade didn't boost performance and as a result it was just 10% faster than the 4770k in this title the margin is reduced it a 7% using the ultra high quality presets so not a big margin but the 87 hurricane was able to add a few extra frames all the same however now at 1440p the 87 or okay was just 2 FPS faster on average it did remain 9% faster for the 1% lower result but still not a great deal in it even with a GTX atti moving on we have the dawn of all three results and here overclocking boosted the 87 arcades performance by 7% at TP using the medium quality preset this meant here it was 13% fast in the 4770k so again not a huge margin by any stretch of the imagination interestingly though increase in the quality settings we see at the 87 arcade was able to extend its lead to 17% and this is as strange as I don't believe there are any settings here that relate to the CPU in any way even so the jump to 1440p saw the overclocked CPUs deliver virtually the same average frame rate while the 87 ROK was 8 percent faster when comparing the minimums now for some Deus Ex mankind divided results and here the 87 Roque was 17% faster in the 4770k with both CPUs overclocked for this test we're using the high quality preset at Tony P and once again the gtx 1080i is handling the rendering work jumping up to the ultra quality preset reduced the 17 percent deficit seen previously - well nothing at all really 1 FPS so within the margin of error it's also worth noting that where is overclocking the 4770k booster performance by 15 percent previously doing so now improve performance by just 7% then finally at 1440 people ran to a severe GPU bottleneck and now all configurations sped out the same 66 FPS minimum and 74 FPS average obviously overclocking either the 4770k or 87 or okay is of no benefit here with the gtx ATI so heavily nailed down moving on we have some dirt for results and for those of you rocking a 144 Hertz display the CPU probably won't matter that much in this tile even stock the 4770k which was the slowest cpu tested allowed for more than 160 FPS to be rendered at all times using the medium quality preset at 1080p overclocked at average 234 frames per second and this meant the 87 RK was 21% faster with an incredible 284 frames per second however with the ultra quality preset applied that margins shrunk dramatically and now the overclocked 87 RK is just 5 percent faster for the average frame rate and slightly less we're looking at the one percent low result then as you might have expected at 1440p use in the ultra quality preset we see the same 124 FPS on average from both the 87 okay and 4770k again under these conditions overclocked in the 87 or okay didn't provide any additional performance moving up we have another racing simulator this time f1 2017 our first that we have the 1080p medium quality results here the overclocked 87 or okay was 18% faster than the overclocked 4770k that's a pretty big margin but of course all the CPUs allowed to GTX 1080i to render well over 100 FPS at all times jumping up to the ultra high quality settings reduce the margin very heavily it has to be said now the 87 okay was just 5% faster than the 4770k when comparing the average frame rate then once again the margin was completely eliminated at 1440p is even the 1080i imposes a strong GPU bottleneck that limits performance across all CPUs tested the last racing simulator that I've tested with is project cars - and here the overclocked 87 ok was 15 percent faster than the 4770k once again we are starting off our testing with the medium quality settings at 1080p which sees a far less GPU limited scenario than the ultra preset that we're about to look at that said though we do see a rare situation where the margin grows when increasing the quality settings and now the 87 okay is 17% fast in the 4770k overall the performance trends are much the same but still it's surprising to say that we're not running into a GPU bound situation not yet anyway using the same ultra quality settings but now the 1440p resolution the 87 eric 8 is just 7% faster and even overclocked the 4770k can't match the stock 7th and 8th generation Core CPUs I assume memory bandwidth for the fourth generation Core i7 processor is the issue here here we have yet another title where the overclocked 8700 K is 18% faster when using the medium quality settings at T P whereas the 4770k managed 228 FPS on average in rainbow 6 siege the 87 era K was good for 269 frames per second interestingly though this is yet another title where even at 4.8 Giga it's the 4770k struggles to keep pace with the stock 7th generation called processors unlike what we saw with project cars to their increase in the quality preset heavily reduces the margin though it has to be said this has been the case with pretty much every other game tested whereas the overclockers america was 18% faster than the 4770k previously it's just 9% faster with the increased quality settings we're also seeing a situation here where the 87 arcade was maxing out the 1080i in its stock out of the box configuration using these quality settings finally at 1440p the 87 RK was just 4% faster on average though it was still 7% faster when comparing the 1% low result that said though while the 87 RK was faster the margin was very slim and with well over 100 FPS at all times the overclocked 4770k certainly delivered a similar experience the final game I've tested with is Warhammer 2 and here we find what almost looks like a frame cap overclocked to the 8700 K and 4770k allowed for the same average frame rate of around 190 fps when using the medium quality preset however if we look at the 1% low result the 87 70 K is 11% faster that said the 4770k did maintain at least 143 FPS using the ultra quality preset while remaining at 1080p reveals a severe GPU bottleneck and now the 4770k an 87 arcade deliver the exact same performance so naturally I expect to see the same thing at 1440p and that is indeed the case 77 FPS on average with a 64 FPS minimum across-the-board alright so what we've seen so far is pretty similar to the original video published a few weeks ago though the addition of the 1440p testing really gives us a clearer look at the GPU limited scenarios let's say before wrapping things up let's have a look at the results across the 11 games tested across all 11 games tested the 80s America was on average 18 percent faster once overclocked when compared to the 4770k if you left both CPUs at their stock out of the box clock speeds the 87 okay would be 25% faster on average and that is of course due the fact that is clocked up to 21% higher still even with the medium type quality settings which alleviates the GPU bottleneck for the most part would you really notice the difference between 128 fps and 152 fps and again those are the 1% low figures for each of the CPUs tested when overclocked so not the average framerate the 1% low figure of course I am testing with the geforce gtx 980ti and i feel that most gamers using such a powerful graphics card would opt for much higher visual quality settings therefore with the ultra quality settings enabled the 87 or okay overclocked was just in 11 percent faster and 18 percent fast with both CPUs stock so not a huge margin therefore modest in the 4770k is 1% lower result of 98 FPS is very respectable then finally at 1440p the overclocked 87 era k was just 4% faster on average though when looking at the 1% lower result it was 9% faster so unsurprisingly the 87 or K is the fastest CPU of the 2 but when it comes to realistic gaming conditions it's really not that much faster so as I said in my previous video I don't think the upgrade from either the 4770k or 4790k to the 87 or k is going to be worth it for the most part gamers won't even notice the difference also please note that both the 4770k and 4790k overclocked 24.0 cue goats will deliver the exact same performance they are based on the same architecture after all although decent gains were seen at times remember we are using a GeForce GTX 1080 I hear you do after you playing at 1080p and you can't be upscaling or making heavy use of anti-aliasing for example meanwhile if you are spending $700 us on a graphics card buy a new CPU or motherboard and memory combo isn't something you really ought to think twice about anyway also as I did say previously you're almost always GPU limited with a core i7 processor even a 4-year old one the updated 1440p results proved this and again I was using a GTX 1080 Ti those using a standard 1080 will see an even smaller margin while those were the 10 70 or perhaps even something slower will see the margin completely eliminated altogether all that said though I am glad I took the time to update the Haswell results of the 4.8 gigahertz overclock with the new memory that I got my hands on and just those extra test conditions and the 1440p testing the next step now though would be to include an unlocked Sandy Bridge processor so a second generation Core processor core i7 processor that is to see how that stacks up it'll be very interesting indeed I have about 20% of the way into that testing with the 2600 ok though I haven't even started do you have a clocking part and my zd68 motherboard just refuses to boot now it's working fine one day shut it down got it the next day to continue on push the power button nothing I've tried a few different things power suppliers and all that and the board refuses to do anything so dead board I will have to get a new Xen 68 board or a newsie 77 board and they're quite hard to track down at a reasonable price but that's my mission for the rest of the week so until I can do that the Sandy Bridge version will be slightly delayed but I flew the next week or two I can get that on the channel anyway that's going to do it for this one if you liked the video please hit the like button that's very helpful leave some comments below also love to hear from you guys and yeah I'll see you again next time or soon yeah I'll see you again soon which will be next time which will be soon because I'll have another video