Gadgetory

welcome back to how our unboxed for a video about high-end CPUs you know the eight core ones that prior to March 2017 nobody could actually afford but now kind of can yeah so previously is in five months ago if you wanted an eight core 16 thread CPU you'd pretty much have to refinance your home the temptation though was just so great that many did of course they all wanted the core i7 69 50 X but with interest rates the way they are they had to settle for the piddly little 69 hundred K so the point I'm trying to make here is that previously Intel's eight core CPUs were stupidly overpriced but now we have the core i7 78 20x and it's less stupidly overpriced the reason you tell us high-end desktop platform pricing has smartened up a little bit is of course due to risin 7 the eight core 16 thread r7 1700 particular was introduced back in march at an unbelievable MSRP of $330 us and if that wasn't hard enough to believe it can be purchased online for his little as $290 us today intel's response to Rison was the core x range and that includes the core i7 78 20x though I'm not quite sure anyone at Intel actually understands how pricing works I mean they obviously get that money is good and that they want more of it so I guess that explains why their new 8 core model comes in at $600 us I guess in their defense they probably think at that price they're basically giving them away okay guys so that's probably the most sarcastic intro of ever done it's at least top 5 material but it's really hard to take this comparison seriously when one CPU has an MSRP of $600 and it's currently retailing for 680 dollars while the other has an MSRP of 330 dollars and is currently selling for $290 and on paper they are quite similar the cheaper CPU can also be paired happily with a $100 motherboard while the more expensive CPU requires a two hundred and $20 plus motherboard so it's pretty clear AMD's Rison seven seventeen hundred is significantly cheaper than Intel's Core i7 78 20x and yet they both pack eight cores with 16 threads what I want to know is just how much better is Intel's offering we've already done a lot of Rison and core X testing here at harbor unboxed so we have a pretty good idea of how these CPUs compare that said I wanted to run some additional tests and initially this video was going to be one of those for science type deals where the information is kind of pointless for average consumer but it looks into something most other benchmarks don't what I wanted to see was how the skylake X based core i7 7820 X compared to the broad well II 6,500 K clock for clock at 4 gigahertz using the same ddr4 2666 memory I then thought well you guys are going to want to see Rison 7 in the mix so I tested the 1700 at 4 gigahertz using ddr4 2666 memory then I thought the AMD fanboys won't like that too much even if we are doing it in the name of science I'll get things like Steve risin scales better than Intel with high frequency memory you're handicapping rising to make Intel look good unsubbed so then I thought in addition to testing rising at 4 gigahertz using ddr4 2666 memory I would also retest using ddr4 3200 memory and then everyone would be happy right no because now we're showing Rison at its maximum overclock using high frequency memory while the 78 20x is limited to 4 gigahertz with 2666 memory so I retested the 78 20x at 4.5 gigahertz using ddr4 3,200 memory with the mesh also overclocked to 3 gigahertz by this point I was heavily sleep-deprived but the testing was done so I moved on to write a coherent and to-the-point introduction you know no more than say a minute looking back I'll admit things got a bit away from me here but the good news is we do now have 19 graphs and 6 CPU configurations to check out so the real fun is just about to begin rather than start with the productivity benchmarks we're going to skip to the dessert for this video we're starting with the game benchmarks consider it means trying to make amends for that intro first up let me explain a bit about what's going on here with all these foreign looking yellow bars the three yellow bars at the bottom represent the horizon 7 1700 core i7 6900 k and core i7 78 20x all clocked at 4 gigahertz using ddr4 2666 memory the two sets of blue bars compare the r7 1700 and I 778 20x again at full you Hertz but this time with ddr4 3,200 memory then the golden bars at the top these are showcase in the i7 78 20x in all of its glory at 4.5 gigahertz using ddr4 3,200 memory with the mesh interconnect overclocked to 3 gigahertz so you got a lot good let's jump into it the results that you've been staring at while I explained my color coding is battlefield 1 and for the graphics card we have the GeForce GTX 1080 Ti looking at the apples to apples results those yellow bars we see that the 1600 K is a few percent faster than the new 78 20x in this title meanwhile the 78 20x is just 13 percent faster than the rise in 7 1700 now what's interesting you note here and we did see these in our 30 game benchmark comparison recently using the risin 5 1600 and core i7 7800 X is that once we up the memory speed the r7 1700 is able to roughly match the 78 20x in fact the increased memory speed really does nothing for the Intel CPU in this game the 78 20x is now just 2 percent faster when comparing the minimum frame rate winding Intel's 8 core CPU right up doesn't really help that much either at 4.5 gigahertz the 78 20x is now just 3% faster than the humble r7 1700 civilization 6 is a game where rise and has in the past looked quite poor in relation to intel's quad core offerings such as the 7700 k and even to this day looks a little bit sluggish however throw the core x-range into the mix and verizon starts to look like a bit of alright clock for clock with the slower 2666 memory Verizon was 11% faster than the 78 20x those shockingly 14% slower than the 60 900k increasing the memory frequency to 3,200 boosted the 78 20x performance by a whopping 17% though less whopping when compared to the i7 1700 s almost 30% performance increase we're only seeing a 20% boost in memory frequency here so the 2666 B must have been imposing a serious bottleneck for risin overclocking 278 20x to 4.5 gigahertz help to extract a further 11 percent performance but even so still trailed the r7 1700 by a 9% margin in our 30 game benchmark Civilization six provided the 7800 X with one of its worst results against the r5 1600 so let's move on to see how things look in f1 2016 while things do look much more competitive here looking at the ddr4 2666 numbers we do yet again file the 1600 K is actually fast in the 7820 x1 match clock for clock the 78 20x was also just 6% fast in the r7 1700 when looking at the minimum frame rate upping the memory frequency to 3200 once again does little to nothing for the 78 20x the gains Verizon weren't huge here either but it was enough to see the r7 1700 roughly match the new Intel 8 core CPU that said squeezing a bit more from the 78 20x did place it firmly ahead of the r7 1700 by around a 6% margin moving on with our Far Cry primal which might seem like an odd choice given it's not that well optimized for high coil count CPUs but it often delivers interesting results so I thought it was worth a look here we see when using the ddr4 2666 memory that the 78 20x is 13% slower than the 1600 K and that's obviously a pretty significant margin it was still 8% fast in the r7 1700 though at least when looking at the average frame rate moving to ddr4 3200 we see that the faster memory boost the minimum frame rate of the 78 20x by 5 percent margin while the r7 1700 enjoyed a 10% boost the rise in CPU was now 8 percent faster when comparing the minimum frame rate and a few frames faster for the average that said once we give the 7800 X all she's got it's able to boost the minimum frame rate by a further 11% allowing it to once again overtake the r7 1700 next up we have hitman and this is yet another game with the older core i7 600 case slays both the Rison 7 1700 and core i7 78 20x when mass clock for clock the r7 1700 all so makes out quite poorly with the ddr4 2666 memory as it was 10% slower than the 78 20x and around 17% slower than the 600 k when comparing the minimum frame rate increase in the memory speed to 3200 didn't do much for the 78 20x and we've seen this quite a bit already the r7 1700 on the other hand enjoyed a 13% jump in the minimum frame rate and a 15% increase for the average so it's not far behind the 78 20x now that said though turning up the heat further with the core i7 78 20x at 4.5 gigahertz helped to improve performance by a further 8% and now the Intel CPU is comfortably faster than the r7 1700 finally we have total war Warhammer and this is a title that has been well optimized to take advantage of Rison through a few handy updates as you can see even with ddr4 2666 memory the rise in CPU is able to lay waste to not just the 78 20x but also the 1600 K in this title it's also worth noting that the Broadwell ECP was again fast and then Intel's new 78 20x delivering an 8% greater minimum frame rate upgrading a ddr4 3200 did improve the 7820 axis minimum frame rate result by 10% which is decent though the r7 1700 did see a massive 16 percent performance bumpy in fact even at 4.5 gigahertz with a 3 gigahertz measure overclock the 78 20x was still slightly slower than the r7 1700 in this title moving on from the gaming benchmarks I have a few synthetic and application benchmarks to look at before checking out memory performance and then the power consumption Cinebench r15 is good for measuring raw CPU performance and here memory performance has a little impact interestingly the core i7 78 20x is fast in the 1600 K when match clock for clock albeit just 4% faster when it comes to multi-threaded performance the rise in 717 hundred and 78 20x are on par though the Intel CPU does enjoy a 5% advantage in the single threaded scenarios overclocked to 4.5 gigahertz the 78 20x naturally pulls ahead and can now be seen achieving a score of just over 1,900 points next up we have some compression and decompression performance with 7-zip from what I've gathered rise ins SMT feature helps massively with decompression work but isn't utilized for compression I haven't looked into this properly yet it's just a theory but in the case of Verizon it is world's better at decompression than it is compression which for most uses is fine as most do significantly more decompression work anyway clock for clock the 1600 K and 78 20x are very similar in this test while Verizon was noticeably faster for the decompression test but significantly slower for the compression test interestingly like what we did see quite a few times in the gaming benchmarks increase in the memory speed did help rise in quite a bit but had very little impact on the 78 20x overclocked a 4.5 gigahertz the 78 20x took a good step forwards and is now able to match the decompression performance of Verizon the blender render test is measured in seconds so lower is better here memory frequency also has little to no impact on performance so this didn't help rise and close the gap on the Intel CPUs when using the ddr4 3,200 memory horizon was 8% slow in the 78 20x in this test when compared clock for clock and 23% slower once the 78 20 X was overclocked to the max and it's four point five gigahertz frequency so a pretty solid win for Intel here we find a similar story when testing with Corona the 78 20x is a little bit faster than the 1,600 K and a lot faster than the r7 1700 with both CPU is overclocked to the max the 78 20x was 16% faster of course it does cost more than twice as much so technically not a win in terms of price versus performance before we get to the power consumption figures here's a quick look at the Ida 64 cache and memory benchmark results there are a few graphs to get through but we'll blow through on pretty quickly here we see that rise ins dual channel memory controller is quite limiting in terms of bandwidth compared to the broad welly and Scala hex CPUs which use a quad channel memory controller whereas the 78 20x can push 66 gigabytes per second for the read through put the r7 1700 was limited to just 40 gigabytes per second meanwhile once overclocked the 78 20x hums along to the tune of 81 gigabytes per second which is obviously quite mega what's interesting to note though is that the core i7 1600 K is significantly better than both the r7 770 100x when it comes to memory latency Rosen does improve with higher clocked memory as does the 78 20x but I guess that's to be expected Rison is well down when looking at the level 1 cache throughput it's basically half that of the Intel CPUs that said though well down ballot the latency performance is much the same rise ins level 2 cache performance is excellent smashing the 600 K clock for clock and even beating the 78 20x overclock the 78 20x does pull ahead but even so Verizon is still very strong here although throughput overall was weaker the 1600 Ches level 2 cache offers a very low latency Rison is also quite good heat while the delay for the 78 20x is quite extreme that is until we overclock the CPU when compared to the 60 900k the r7 1700 s level 3 cache throughput looks excellent however when compared to the 78 20x it does look quite weak indeed however throughput isn't everything and here we see that despite the big ballot the latency is still very poor for the 78 20x even once overclocked so in terms of responsiveness Verizon has a big advantage here since memory frequency has little to no impact on the overall system consumption I've only looked at the ddr4 2666 figures here along with the 4.5 gigahertz overclock for the 78 20x clock for clock the core i7 600 K was very efficient pushing total system consumption are just 206 watts in Cinebench r15 multi-threaded The Horizon 7 1700 didn't do quite well drawing just 248 watts while the 78 20x was quite a bit hungrier heading 268 watts however it was once overclocked the 78 20x increased total system consumption by 26% and is now seen consuming 36% more power than the r7 1700 a pretty significant figure indeed well then there's some very interesting things to discuss here let's start with the core i7 1600 K and 78 20x comparison was quite shocking to find that were compared clock for clock using the same speed memory that the oldest 600 K was faster in every single game we tested at times was significantly fast a lot what we saw with Civilization 6 even giving the 78 20x an advantage of the faster ddr4 3,200 memory which mr. okay doesn't actually support it was rare that the sky like axe CPU would hit the lead even overclocked to 4.5 gigahertz with the three gigahertz mesh frequency overclock this 78 20x was only able to match the 600 K in most of the titles tested and realistically we could go ahead in the squeeze a few hundred megahertz more out of the Broadway Lee CPU adding insult to injury is the fact that the 78 20 X consumes significantly more power to deliver similar performance of the previous generation part for that 4.5 gigahertz overclock you will require a seriously high-end liquid-cooled solution if you want to avoid heavy throttling when it came to application performance the 78 20 X did look much better though even then it wasn't always clearly superior to the 16 hour okay for example we saw similar performance in the 7-zip application while the Cinebench r15 numbers weren't drastically different either the 78 20x was marginally better in the blender and corona tests but not to the degree where you'd find yourself that excited about the results the only advantage of the 78 20x really has over the six 900k is the fact that it's cheaper around 35% cheaper which is obviously a big deal that said if you gave me the choice of either of these CPUs at the same price I'd probably take the 1600 K moving on to the 78 20x versus r7 1700 comparison it's pretty clear that in terms of value the rise in seven CPU is in a different league I really don't think even the most loyal Intel fanboy could argue otherwise and as always I'm not sure why they'd bother anyway once you factor in things like the motherboard cost to the equation the 78 20 X costs around a hundred and thirty percent more than the r7 1700 needless to say it was never anywhere near that much faster comparing the maximum overclocked gaming performance results really was pretty similar experience overall the 78 20x for example it enjoyed a decent winning hitman though the r7 1700 was noticeably better in Civilization 6 and yeah as I said the rest was really much of a muchness the 78 20x was 23% faster in blender and 16% faster for the corona benchmark so it was hands-down faster for these productivity workloads but yeah keep in mind it wasn't 130 percent faster and for that extra performance it did consume 36 percent more power so also keep that in mind in the end I feel like the sky like X architecture is it best a side step from Broadway Lee and this has a forwarded risin a great deal of breathing room hadn't L been the slightest bit aggressive with their pricing then risin would have a serious fight on its hands but as it stands at 600 ish dollars you'd be mad to spend that kind of money on the 78th 20 X where you can get comparable performance for less than half that price and that's where I'm gonna leave this comparison if you like this video we'll go ahead and make it official even share their video and of course don't forget to subscribe as I do have a heat more benchmark content coming up on the channel very soon I'm your host Steve see you again soon guys you