Gadgetory

back when I first tested the sky based core i7 6700 K a few weeks prior to its release I ended up publishing results that were on par with the 4790k and at times the 6700 K was even slower there are a few reasons for why I found these results unaware at the time my a driver's ed 97 motherboard was being overly aggressive with a maximum turbo frequency or at least how many cores would reach said frequency on the 4790k in other words it wasn't adhering to the Intel spec the other and let's say primary reason was down to the fact that the 6700 K really isn't any faster than the 4790k for the most part IPC performance has only been improved by up to 10% leading tech Sydenham tech also found in the clock for clock gaming comparison sky coz no faster than as well and in many instances was even slower which was an odd discovery in the end I concluded the intel's most powerful and efficient architecture yet sky lake was for the most part a disappointment obviously at the time this wasn't even taking into account the fact that six months later the 6700 K would still retail for around 20 percent more than the 4790k the continued incremental baby steps entails been taking for years now made me question if there was any real need for Sandy Bridge core i5 2500 K and core i7 2670 gaming due to my early adopter syndrome my 2600 cases still was retired in favor of the 3770k which was of course retired from 4770k and so on however after looking at a number of gaming performance articles recently focusing on games such as rise of the Tomb Raider and Tom Clancy's the division for example I noticed the 2600 K in 6700 K weren't that far apart especially considering these articles one overclocking the CPUs putting the 2600 K to ten o'clock disadvantage interested to know how not only the core i7 6700 K compares to the five-year-old 2600 K but also the 3770k 4770k and even 5775 C in a clock for clock comparison I resurrected my old gaming rigs the objective is simple to compare five generations of keio 7000 suspicions this meant clocking and locking them all at four gigahertz with any frequency boosting technologies disabled along with feed step the process is using ddr3 memory target 2400 megahertz and the 2600 K was the only ddr3 processor to drop down to 2133 megahertz as the memory controller isn't stable at 2400 megahertz meanwhile the ddr4 memory used by the 6700 K was clocked at 26 66 megahertz which we found previously to be about the sweet spot initially I had planned to only test them using a flagship graphics cards such as the geforce gtx 980ti having done so it became clear that in conjunction with the lesser card such as the GTX 970 we could paint a more complete picture so it was settled the five core i7 processors clocked at four Giga her to be tested in nine cutting-edge games at 1080p using the gtx 980ti and gtx 970 along with the game data I've also included power consumption figures and raw memory bandwidth data that being said let's move on to the benchmarks the memory bandwidth figures between the processor and the DDR memory range from 30 gigabytes per second to a little over 35 gigabytes per second the ddr4 2666 enabled sky processor providing the best result that said the Broadwell and Haswell processors weren't far behind using ddr3 2400 memory it is interesting to note that the l3 cache yes cache since it's not a currency is considerably higher on the skylight processor whereas the previous four generations all delivered similar performance using a gtx 980ti graphics card we find there's virtually no difference between the core i7 2660 700k model the sky processor was just 4% faster when comparing average frame rate and 2% faster for the minimum frame rate swapping out the 980ti for the more popular GTX 970 you reduced the margin to just a single percent when comparing the average frame rate battlefield for single-player campaign isn't particularly - actually on the CPU so it'd be interesting to see the correlation between the gtx 980ti and GTX 970 on these processes in more demanding titles here we find some interesting results when using black ops 3 that has well Broadwell and scaling processes all delivered similar results however when we move from Haswell to Ivy Bridge we see a 9% reduction in frame rate followed by a further 5% reduction when moving for Ivy Bridge to Sandy Bridge this means overall the 2600 K was 15% slower than the 6700 K when comparing minimum frame rate though it was just 5% slower if you look at the average frame rate now this is very interesting reducing the available GPU power with the GTX 970 eliminates any difference between the 2600 K and 6700 K completely this still very powerful graphics card doesn't have enough rendering power to separate these processes in black ops 3 all out poor might not be the best example given the game strange tendencies with memory bandwidth in certain areas of the game whereas the 2600 K through to the 5775 see all deliver similar performance the 6700 K is 11% faster than the broad well and has well processes while beating the Sandy Bridge 2600 K by 17 percent margin this just shows what a strange animal fallout 4 is here we see that despite using a much slower graphics card we receive very much the same performance this clearly indicates a serious cpu bottleneck in this game which is shocking given the average frame rate for most of these processes is below 60 fps GTA 5 is probably the best game we could use for such a test so these results are worth paying attention to like black ops 3 with the 980ti we find the Haswell Broadwell and skying processes all deliver similar performance with just a few frames per second separating them however if we look at the Ivy Bridge 3770k we see a noticeable decline in performance and then again when moving to 2600 K the average frame rate of the 2600 K was 16% slower than the 6700 K while the minimum frame rate was 22% lower that said it is important to know that the 2600 K did manage minimum frame rate of no less than 90 fps so still obviously very playable this is again where things get very interesting island with the still very capable gtx 970 a $300 graphics card you see very little difference between these processes the 6700 K is now just 5 FPS or 6% faster than the 2600 K and just 3% faster than the 3770k for a single player game Just Cause 3 is very CPU intensive we see a bit of that in these results even so the 6700 K doesn't exactly crush the 2600 K as it only provided a 6% greater average frame rate coupled with an 8% better minimum frame rate those are some pretty easy difficut results for the gtx 980ti handling the rana rig work at 1080p as we've seen numerous times already using a more commonly featured GPU in gaming systems the difference between these various processes becomes even smaller the 6700 K is now just 5% faster than the 2600 km and comparing the minimum frame rate and even less when looking at the average the popular tactical shooter Rainbow six siege plays the same regardless of whether you're running a core i7 2670 K the 6700 K was less than 2% faster when comparing the minimum frame rates and nothing more needs to be said here give him what we saw in testing Rainbow six siege with the 980 TI we didn't expect to see anything different with the gtx 970m as expected we didn't as is the case with all multiplayer games it is impossible to accurately test this portion of the game forcing us to benchmark a single player training mission as you can see with the gtx 980ti crushing pixels there's virtually no difference between the 2600 K and the 6700 K the margins are reduced further with the GTX 970 and here the 6700 K was just 5% faster than the 2600 K when comparing minimum frame rates rise of the Tomb Raider is a visually stunning game that's also fairly CPU intensive that said when armed with the gtx 980ti all five processes array would deliver similar performance 1080p60 700k was just 5 percent faster than the 2600 came frame rate now with the 972 700k is just a single frame faster than 2,600 K while the average frame rate is identical across the board which are three results are surprising given we expected to see similar average frame rate results with a rather large variation in the minimum frame rates it appears the opposite is true here the 2600 K never dipped below 55 FPS making the 6700 K just 7% faster however if we looked at the average frame rate the 6700 K was 13% faster again not a huge difference but more than we were expecting given the minimums finally as we've discovered time and time again when swapping out the 980ti for the 970 we see any previous margins dissolved into almost nothing in fact this time was absolutely nothing as the 2600 K is 6700 K deliver the same minimum frame rate result this is really where Intel has made the greatest strides over the past few years efficiency despite being fast in the Sandy Bridge and CPU bound games the sky architecture is seen consume 14% less power in just 4:3 granted going from a total system consumption of 335 watts down of 287 watts isn't game-changing this improvement still comes while achieving greater performance testing with rise of the Tomb Raider which is similar power consumption figures here the 6700 K consumes slightly less power than both the 5775 C and 4770k okay so let's get right down to making sense of all that data with a few fun line graphs what you're looking at here is the averages for those minimum frame rates across all nine games tested the light blue bar represents the gtx 980ti data with the darker blue bar the GTX 970 while the corresponding processors can be seen below here we see there's just a 2% performance difference between the 2600 K 6700 K using the 970 that margin is increased to 9 percent in favor of the 6700 K once we increase the rendering power with the 980ti honestly these margins aren't significant at all however you might be sitting there saying come on man stop shaping the results with those non CPU intensive aimes well I hear you so here's a 5-game average using the witcher 3 black ops 3 fallout 4 Just Cause 3 and GTA 5 what's interesting to note here is that again with the GTX 970 the 6700 K is only marginally faster than the 2600 K just 3% faster in fact the 980ti on the other hand does allow the 6700 K as well as the 5775 C and 4770k to stretch their legs a bit as a result the 6700 K was 17% faster than the 2600 K with an 81 FPS average minimum opposed to 69 FPS for the 2600 K now there are a few things to consider here if we were to test all five processes clocked at 4.5 gigahertz for example the gtx 980ti margins would likely become closed out if only slightly likewise if we down clock the processors to 3.5 gigahertz the margins might grow slightly too naturally the margins would also grow when using the more powerful GPU configurations particularly if SLI or crossfire is employed for example still considering all these factors I believe it's fair to say the owners of an unclogged Sandy Bridge processor really have no immediate need to upgrade this is particularly true when spending $300 or less on a graphics card if you invested around $300 on a Clio 720 600 K 5 years ago you're probably pretty happy with the run it's had especially for the kind of gamer that Springs from a mid-range GPU once every year or two be sure to let me know what you think about the results are you surprised about how well the Sandy Bridge 2600 K handles the GeForce GTX 970 I know I was having initially tested with the gtx 980ti don't forget to hit like and hit subscribe on your way out and I'll see you guys again next time you