Gadgetory

well a lot of you guys asked for it so I've been working hard to deliver it I'm of course speaking about a core i5 version of February's five generations of core i7 processors video which took an in-depth look at gaming performance on every mainstream core i7 processor since 2012 Sandy Bridge 2600 K for this video I've gone one better and included every generation of Core i5 processors released to date with the exception of the dual core clarkdale processors this means 2009 Linfield Core i5 750 has been included along with Sandy Bridge Ivy Bridge has well Broadwell and of course skylight core i5 processors the format for this video will be much the same as the core i7 version all processors have been locked at 4 gigahertz with any power saving or throttling features disabled that said I have tweaked the benchmarks ever so slightly some of my most beloved critics pointed out that in the previous core i7 comparison only GPU bound games we use for testing obviously that wasn't true given the inclusion of The Witcher 3 fallout 4 Just Cause 3 and Grand Theft Auto 5 still of the nine games included four of them were mostly GPU bound so you might ask why include them in its CPU comparison the reason is simple the entire point of the comparisons are trying to determine if game is still rocking a Sandy Bridge processor really need to upgrade to Intel's latest skylake architecture testing the CPU bound games doesn't really paint an accurate or complete picture especially when the majority of PC games are very much GPU bound games addicted to Star Wars Battlefront for example receive the same performance with either the core i7 2670 K even when running extreme GPU setups whilst the opposite is true in any of the CPU bound games I just mentioned still in an effort to try and keep everyone happy I have included ARMA 3 this time ridiculously CPU bound open-world military tactical shooter also included is civilization beyond Earth and ashes of the singularity which was tested both in the DirectX 11 and DirectX 12 modes there's a heap of data in this video so for those that want to quickly check out all the graphs it might be best to check the link in the description to the text-based version at our website I'm starting with Star Wars Battlefront a game which we already know isn't very CPU demanding at least in the single-player tutorial mode where I have to test for accurate results this is a great indicator of typical game player where the GPU is the limiting factor here we see that even the lowly old 760 is able to deliver strong performance with the gtx 980ti as expected the GTX 970 results are much the same only the 760 falls behind the pack though not by a significant margin battlefront was also tested using a pair of Radeon r9 390x graphics cards in crossfire here we see that while the all-important minimum frame rates are surprisingly close the more recently released core i5 models delivered better top-end performance like battlefront battlefield 4 is another popular title that at least when playing the single-player mode isn't very CPU demanding here are all six processors including the 760 provided competitive results with the gtx 980ti similar performance trends were seen when downgrading the GPU to the GTX 970 and it's only the 760 that falls a few frames behind the newer Core i5 models now with a pair of 390x graphics cards we see that there's a slight difference in performance between most of the Core i5 processors still it's important to note that even the 760 was able to push past the 120 FPS barrier without dipping below 100 FPS at any time which is very surprising at least to me here we have yet another mostly CPU bound first-person shooter that even the old core i5 760 still plays well with and all these results with the gtx 980ti both impressive and surprising as you'd expect downgrading the GPU to the gtx 970 only reduce the performance margins and here we see the skylake 6600 k beating the old 760 by just 5 FPS which equates to a mere 7% performance margin increasing the rendering firepower with to 390 X graphics cards only solve for FPS separate the minimum frame rate of the 760 and the 6600 K civilization beyond Earth places a fair load on the CPU though the results we see when testing with the gtx 980ti just 1080p aren't extreme the 6600 K is just temps and faster than the Sandy Bridge Base 2500 K that said the Broadwell 56 75 C provided the best result with a minimum of 84 fps and a 150 the FPS average moving to the GTX 970 reduced the margin between the Haswell Broadwell and sky processors they can all be seen dipping to just 80 FPS with a similar average result the Linfield 760 really bombed out in this test dropping it was 55 FPS making it 31 percent slower than the Haswell part and 25 percent slower than the 2500 k that came after it I should point out that there was no 390x crossfire testing for this title since crossfire wasn't working this is the case for quite a few of the games we tested with unfortunately things get crazy in our fallout 4 boston city tests and here games require some serious cpu firepower the 56 75 c comes out ahead here as i suspect it's able to put that massive l4 cache to good use we see a nice progression in performance from Sandy Bridge to Ivy Bridge and then to Haswell which is nice the core i5 760 really struggling this title with frame dips as low as 33 FPS because the section of Fallout 4 that we benchmark is so extremely CPU bound downgrading to the GTX 970 has a very limited impact on performance Grand Theft Auto 5 is one of the best examples we have of a well balanced game and it does a good job of taking advantage of the latest hardware that's said with the gtx 980ti handling the rendering the 4690k 56 75 C and 6600 K were virtually indistinguishable the performance did start to drop off with the 35 70 K and was reduced further by the 2500 K although the frame rates were still very high the 760 was good but slower than the more modern Core i5 processors interestingly the core i5 760 recovered somewhat with the GTX 970 was able to match the 2500 K still the 35 70 K was a decent step up from the 2500 K and the 4690k provided yet another step forward before we hit a brick wall with the new abroad well and skylake processors Just Cause 3 is a super fun and very CPU intensive game that said even with the gtx 980ti gaming at just 1080p the Sandy Bridge 2500 K does very well the same unfortunately can't be said for the 760 which dropped almost 30% lower than the 2500 K of 48 fps even with the slower GTX 970 handling the rendering the 760 still looks pretty slow it allowed the framerate to hit 47 fps at times though admittedly the average framerate isn't that bad Just Cause 3 is another title doesn't support crossfire so we skip testing the three 90s graphics cards here as well Rainbow six seed isn't very CPU intensive at all and as a result the core i5 760 is able to hang in there with the much newer core i5 processors the 760 kept the minimum frame rate above 120 fps so obviously game is very playable on this processor the GTX 970 provided similar performance margins and although the 760 was the only CPU to dip below 100 FPS at 92 fps the game was still very playable rise of the Tomb Raider was a game that I initially thought it was quite CPU intensive when first played as it does make quite good use of at least four threads however as you can see even the 760 is able to hang with the much newer i5 processors such as the 6600 K for example GTX 970 so the 76020 500 K and 35 70 K all delivered the exact same performance while the newer Core i5 processors were just a frame or two faster testing for The Witcher 3 Wild Hunt took place at the town of novigrad here the game is quite CPU demanding despite that the performance margin between the 2500 K and 6600 K was almost non-existent it was only the 760 that dropped off in performance albeit by just 5 fps when comparing the minimum frame rates the GTX 970 provided some interesting results and the 760 was able to match the 2500 K here which meant essentially it wasn't really any slower than the newer Core i5 processors which is another really big surprise when we move to the 398 crossfire configuration the 760 does start to look dated and although the minimum frame rate here isn't that poor the average frame rate would suggest the 760 spent quite a bit of time rendering less than 60 FPS now for the game many of you have no doubt been waiting for ARMA 3 this game puts a serious stranglehold on the CPU and this can be seen when testing the gtx 980ti although the 6600 K 56 75 C and 4690k all provided similar performance there was a huge drop in framerate when moving to the 35 70 K from the 4690k to 35 70 K the minimum frame rate dropped by 15% and we saw a further eight percent drop when going from the 35 70 K to the 2500 K finally we bottomed out a 50 9 FPS with the 760 making a 30% slower than the sky like an Haswell processors now with the GTX 970 we fall just short of 80 FPS with the higher-end Core i5 processors while they saw a significant reduction in performance we're moving from the gtx 980ti to the 970 the same wasn't true for the older 2500 K and 760 in fact those processors drop just one to two frames which meant the 760 was just 17 percent slower than the Haswell Broadwell and skylake processors the best example we currently have of DirectX 12 performance is ashes of the singularity in my opinion so I've added some DirectX 11 and DirectX 12 comparison testing with all six Core i5 processors using the gtx 980ti gtx 970 and fury X using the gtx 980ti we find some interesting results the more modern more powerful Core i5 processors were actually faster when using DirectX 11 as in videos gtx 980ti is still more efficient here due to its hardware design and more efficient drivers as the cpu performance is reduced we see the 35 70 K provides the same performance in DirectX 11 and DirectX 12 by the time we hit the Sandy Bridge 2500 K the processor efficiency has dropped to a level where the lower CPU overhead of the DirectX 12 API overcomes any shortcomings of the gtx 980ti as DirectX 12 performance in this title the 760 amplifier's is further as now the gameplay is significantly better in DirectX 12 mode due to the fact that CPU isn't holding the GPU up as much as we've found more often than not the slower GTX 970 evens up the playing field though again the 760 is faster and running in DirectX 12 looking at the DirectX 12 results the fury X performs exceptionally well and the 2500 K through the 6600 K are able to get the absolute maximum performance from this high-end graphics card DirectX 12 also helps out the core i5 760 significantly and although it's still much slower than even the 2500 K it does provide very playable 50 fps first up we have the power consumption results from tomb raider using the gtx 970 and as you can see there have been improvements along the way with the exception of Broadwell and its power-hungry l4 cache skylake system continued apes and less power than the Haswell system a 19% less than the Sandy Bridge system these note where the improvements that they'll probably fail to blow the socks off the average gamer the cpu-bound just course or a similar 20% reduction in power usage when going from the 2500 K to the 6600 cave interestingly the heavily overclocked Linfield 760 at 4 gigahertz only consumed an extra watt when compared to the 2500 k finally before wrapping things up I've provided some memory and cache performance for those interested Sandy Bridge was tested with ddr3 2133 memory this was the high frequency we could maintain 100% stability with the 760 was limited to 18 33 megahertz well if you've made it this far congratulations and if you just hit conclusion in the video index then shame on you before I break down the results let me just say this video isn't meant to disappoint or upset those who invested in one of Intel's latest guy processes hell I've knowingly purchased a few of them myself after all if I were building a new computer today at the top of my wish list would be a sky core i5 or i7 processor instead the point of this video is to provide those that are still rocky an older generation processor a clear indication of the kind of performance they stand to gain by upgrading going all the way back to the first core i5 process is based on the Linfield architecture we see that Schaick offers a very real performance improvement and even Sandy Bridge was a decent step forward that's there for the most part when paired with a relatively high end GPU such as the GTX 970 the core i5 760 was able to deliver very playable performance in most of the games tested it's really only the extreme titles such as fallout 4 Just Cause 3 and ARMA 3 where there's extra frames do make a noticeable difference of course out of the box the core i5 760 is going to be significantly slower than what was seen here as it operates at just 2.8 gigahertz so be sure to keep that in mind the same can be said for the 3.3 gigahertz though the fully unlocked clock multiplier makes that a dead easy CPU to overclock performance wise the margins are very similar to those found when testing the core i7 processors if we look at the CPU intensive titles only here's what we find using the GTX 970 the skylake and Broadwell processes are fourteen percent faster than the Sandy Bridge 2500 K and eight percent faster than 35 70 K they were however 25 percent faster than the 760 that is a huge game given the seven year age gap with the faster 980ti handling the rendering the sky leak and has all processes were again just 14% faster than the Sandy Bridge 2500 K that said they were now 35 percent faster than the 760 naturally the margins will grow when using more powerful GPU configurations particularly if SLI crossfire is employed for example now if we look at the non CPU intensive titles we find very different results using it the gtx 980ti or GTX 970 the top tied CPUs such as the 6600 K 56 75 C and 4690k were all well under 10% faster than the 760 while they provided virtually the same performance as the 2500 K so depending on what games you play it is possible to make do with a very old Core i5 processor in today's games running a respectable GPU that said when the CPU is called to action at times the 760 looked a little slow though admittedly that didn't happen often still I could easily justify the upgrade to Haswell Broadwell or sky from Linfield I probably couldn't say the same for Sandy Bridge and certainly not for Ivy Bridge DirectX 12 isn't going to bury these older generation processors either that was something a few viewers were arguing in the core i7 video ashes of the singularity which is a very well-put-together title and is extremely CPU intensive shows how DirectX 12 doesn't hurt slower processes but rather lens of a helping hand the gtx 980ti and fury x results are a perfect example of this looking at the fury x results which are the most extreme we said that although Sandy Bridge folds away slightly under DirectX 11 that isn't the case when using DirectX 12 here it's able to match the performance of the Haswell and sky chips although DirectX 12 helps much older Core i5 760 it wasn't able to completely make up the performance deficit there are obviously still similar imitations here that can't be overcome by the low-level API that's hardly surprising so for now it appears as though DirectX 12 is only going to prolong the life of processes such as the 2500 K rather than ended virtual reality is another hot topic and I was told a few times in regards to the core i7 video that 2600 K would struggle with VR as a minimum of 90 FPS as desired I don't know about you but a mere 12% drop in performance from the 6600 K to the 2500 cable looking solely at CPU intensive games tell me the 2500 K should be VR ready so for now the gaming future looks bright for anyone rocking a Sandy Bridge Core i5 or newer and I suspect that too good many of you thanks for joining me again for another episode of hardware unbox I'm your host Matt as always and I look forward to seeing you guys next time