Gadgetory

welcome back to how our unboxed we finally have the missing pieces of the sky like x-series after hastily announcing the 18 core part back in late May during the Computex 2017 trade show we now have it in hand four months later I don't doubt that Intel always plan to release the sky like x-series this year but did they plan to offer say higher than a 12 core part before they caught wind of ambe's thread Ripper it also seems unlikely that they plan to announce the x-gen ion platform in late May and then release at the following month and this was evident by the heavily rush to motherboard development and just the mess that ensued so in late June we got the first core iron iron part the 10 core 70 100x along with a few core i7 models offering 86 and even for cos the 70 100x is a $1000 part but we knew there was also going to be a $1200 12 core $1,400 14 core $1700 16 core and a $2,000 18 core in early August Intel announced the official specifications for the higher core count parts and we learned that the 18 core and 16 core models would feature fairly low base clocks though this was to be expected before we cover the i9 79 80 XC and 79 60 X in detail it's certainly well worth mentioning that during all the chaos of Intel's latest desktop platform release AMD released their rise and thread Ripper series on August 10th the thread Ripper 1950 X in 1920 X arrived and with that we received a $1000 16 core CPU that shamed Intel's then flagship 7900 X of course as we're about to see there was more to come from Intel but even so at the time we questioned if a $1200 12 course 79 20 X could even match let alone beat the 1950 X and we suspected that would probably take the $1400 14 core 79 40 X to match AMD 16 core CPU well unfortunately we don't have the 12 call all the 14 core models to find out instead intel served up the ultra expensive 16 and 18 core model for this review that said though this will give us a very good idea of whether twelve and fourteen core version will slot in and of course we're always keen to see what the flagship parts have to offer so to quickly recap the core I 979 80 XE packs 18 cores and with a type of hurting support offers and insane 36 threads each core has its own 1 megabyte level 2 cache and this means there's a total of 18 megabytes of cache meanwhile there's also 24.7 5 megabytes worth of l3 cache and due to the restructuring of the cache for the sky like ex architecture the 79-80 XE only has slightly more l3 than the previous generation 69 50 X the 79 80 XE operates at a base frequency of 2.6 gigahertz with a Turbo Boost 2.0 frequency of 4.2 gigahertz and a turbo boost max 3.0 frequency of 4.4 gigahertz the 16 core 79 60 X features a base clock speed of 2.8 yoghurts with the same turbo boost frequencies it also has a slightly smaller 22 megabyte l3 cache well there's a total of 16 megabytes for the l2 both CPUs offer the full 44 PCIe lanes quad channel memory support and a 165 watt TDP rating though that routing pretty much means nothing at this point both CPUs use the LGA 2066 socket and are supported by existing X 299 motherboards for testing I'm using the gigabyte Oris X 299 gaming 9 with 32 gigabytes of g.skill stridency RGB ddr4 memory clocked at 1600 megahertz so ddr4 3200 using cell 14 timings the GTX 1082 I was used for all the testing while we've thrown vega 64 liquid-cooled in for a few of the game tests the focus though will of course be on productivity testing this is where the a-tingle and 16 core CPUs make the most sense after all all right so let's get into the benchmarks first up let's check out the memory bandwidth performance these ddr4 quad-channel memory controllers look to be good for around 62 to 64 gigabytes per second of memory bandwidth using ddr4 3200 memory please note all configurations were tested using the same a ddr4 3200 CL 14 memory Cinebench r15 s multi-threaded test gives us a very good idea of how the productivity tests are going to look assuming that all cores can be fully utilized here we see the 79 60 X is just 4% faster than the 1950 X for the multi-threaded workload both are of course 16 core parts the 79 80 XE is 10% faster than the 1950 X with an impressive score of 3317 points though it does cost twice as much at $2,000 and we'll certainly discuss that more towards the end of the review of course when it comes to single thread performance the higher clock sky like X parts are able to best thread Ripper before we move on to some more serious benchmarks I just wanted to see how these extreme CPU stand up in PC mark 10 which looks at office type workloads which aren't typically call heavy to my surprise thread Ripper did score very well in this test and did considerably better than the sky X CPUs for example the 79-80 XE looked particularly weak here of course though none of these CPUs are by any means slow for these general office type tasks so it's somewhat irrelevant excel is an office type benchmark though which can utilize many threads especially when running the extreme Monte Carlo simulation here the thread Rupa 1950 XR really impressed taking just one point 66 seconds however the 16 and 18 course Kollek X parts go one better as they dip down into the 1.4 second range allowing the 18 core part to complete the workload 14% faster next up we have veracrypt for those of you unaware veracrypt is an open source utility used for on-the-fly encryption and features optimized implementations of cryptographic hash functions and ciphers which do boost performance of modern CPUs it also supports paralyzed encryption for better multi-core systems as well as harbours accelerated aes to further improve performance in short it takes full advantage of the many cores these CPUs have on offer and we see that here that said though the thread Ripper 1950 X did provide the best results and even the 79-80 XE couldn't beat it so this is a very disappointing result for Intel next that we have some non encrypted compression and decompression performance using 7-zip whereas hyper threading sees a similar level of efficiency when compressing and decompressing SMT is significantly more efficient for decompression work for the decompression worker thread rippin 1950 X is again able to edge out the 79 XE though it was much slower for the compression test so if you do a lot of compression work then it looks like the Intel hike or count CPUs are king here handbrake is a popular application for video encoding and we've used it to convert this 4k h.264 video to 1080p using h.265 and recorded the average frame rate here the 79 60 X 80 managed 22 point 3 FPS while the 79 60 X was actually 8 percent faster the reason for this is down to utilization handbrake really only utilized about 70 percent of the 18 core CPU and because the 79 80 XE has a lower operating frequency the 79 60 X actually came out on top for me the Premiere Pro CC results are of the most interest as a program I use on a regular basis and for the moment I'm rendering 4k videos on my core i7 69 50x editing machine I do plan to upgrade to thread Ripper soon but I have to say the 79 80 XE and 79 60 export times are enticing that said that will be interesting to see how many people can justify spending 70 percent more on the 79 60 X to shave off around 13% from the incurred time it's also worth noting that once again the 16 called 79 60 X was faster than the 79 80 XE due to the fact that premiere isn't able to fully utilize the 18 core CPU moving on to the rendering tests we have blender and for this first test we are using the rising graphic 27 benchmark here the 79 80 XE and the 79 60 X both deliver similar results taking about 12 seconds each and this meant at best they were 10% fast in the thread Ripper 1950 X the gooseberry workload takes some serious firepower to complete in around half an hour and we see this as the 1950 X turns in a time of almost 32 minutes that said though the 79 80 XE and the 79 60 X were mighty impressive here taking around 23 minutes once again the 16 core 79 60 X did it paste the 18 course 79 80 XE the 16 core part was a whopping 27 percent faster than the 1950 X for this extreme blender workload so that was impressive to see Corona comes as a standalone benchmark it renders a fixed scene six times and we take the time it takes to complete the entire task this application apps will loves threads the more the merrier and the quicker you'll be done you can move on to something else here the 79 60 X was 14% fast in the 1950 X while the 79 80 XE was 20% faster now this is one of the better results we've seen for Intel still a 20% increase in performance for a 100% increase in price might be a tough sell povery is another raytracer it's been around for many years now and we're using the official benchmark here the 79 80 XC was just 11% fast in the 1950 X while the 79 60 X was 7% faster so while Intel's new 16 core 1:18 core parts are certainly faster they're not really that much faster all right so it's time for a few quick gaming benchmarks before we look at the power consumption and temperatures here we're testing battlefield one using the DirectX 12 API with the ultra quality preset enabled all the games have been tested using the GTX 1080i and Vega 64 GPUs and first up we have the GeForce results here we see that 3 Ripper doesn't look particularly impressive but it has to be said with over 100 FPS at all times the game was incredibly smooth please note that I'm not bothering the test the thread Ripper CPUs with that game mode enabled resetting your workstation to play a couple of games isn't particularly realistic so I'm only going to test with the distributed memory mode as for the 79 60 X and 79 80 X 8 they do quite well despite trailing the lower core count parts but that was pretty much to be expected then as we change things up by dropping in Vegas 64 liquid-cooled the frame rates don't change drastically for the quad core Core i7 77 40 X when compared to what we saw with the GTX 1080i however the AMD processors perform much better and we're now seeing significantly more competitive performance across the board with Vega so that's very interesting to see moving on we have ashes of the singularity escalation again we're testing with the 1080 TI first here the core ARM processors deliver the best results while thread repeat is comparable to the core i7 78 20x however when we retest with Vega we do see much more competitive results from all tested CPUs the 1950 X again matches the 78 20x but they're much closer to the core i9 models now next that we have civilization six and here the risin seven 1800 X is very competitive with the core i7 77 40 X when using the gtx 1080i meanwhile the 1950 X matched the r7 1700 and both were much faster than the Intel skylake X CPUs particularly the new 16 and 18 core parts switching over to the Vega 64 liquid core graphics card we once again find quite different results and this time of the AMD CPUs are able to pull much further ahead of these Scarlett X parts and even the 7740 X falls behind the thread rippen 1950 ex finishing up the gaming benchmarks we have f1 2017 and this is the only DirectX 11 title we're testing with here the core I 979 60 X and 79 80 X look very weak though keep in mind they're still pushing over 100 FPS at all times meanwhile it's worth noting that the 1950 X offers considerably better minimum frame rate performance than even the 7900 X that said it's worth noting that the quad core 77 40 X is the real hero here pushing well over 200 FPS for the most part switching over to Vega doesn't change too much here though the average frame rate of the 1950 X is now much better than any of the Scarlett X parts and in fact we do say that Rison is the king of the big core count CPUs here next that we have the power consumption figures and it's important to measure using software that stresses all the cores so for that I've used Corona and that works very well providing very accurate results so these load figures are based on the current and benchmark after a single pass I'm reporting the maximum log to result what you see he is total system dry I'm using the Quebec palmate to measure from the wall the third River 1950 X sucked down 275 watts and that meant that the total system consumption was 8% lower than that of the core I nine 7900 x interestingly the 79 ATX he actually drew less power than the 79 60 X in this test both CPUs delivered very similar performance so this suggests that the 79 60 X had to clock much more aggressively to achieve that result and this meant greater power draw under full load the 79 60 X system was consuming around 25 percent more power than the 1950 X while the 79 80 XE consumes 16% more power okay so time to do a little tinkering the 79 80 XE over clocked all calls to 4.1 gigahertz quite easily it's probably possible to go further but for now this is where I stopped due to power drawing a few other reasons the 79 60 X was happy at 4.3 gigahertz and of course previously we got the 70 100x 24.7 year gets thread Ripper has also been thrown into the mix clocked at 4 gigahertz clocked at 4.1 gig 'it's the 79 80 XE produced a multi thread score of three thousand nine hundred and seventy four points so expect to see some people breaking the 4000 barrier with this one that's a 20 percent increase over the out of the box score as well the 79 60 X saw a 17 percent boost as it hit an impressive three thousand six hundred and eighty-one points the thread Ripper 1950 X only managed to boost its out-of-the-box performance by 13 percent but even so that allowed for a score of three thousand four hundred and twenty four points overclock the 79 80 XE was just 16% fast in the 1950 x so it'll be interesting to see what kind of difference this makes in the real world let's check out our extreme blender workload well honestly it's not really a huge amount both the skylark X parts now 9 percent faster while thread Ripper is now at 11% faster even so the 16 core and 18 core Intel CPUs are blistering fast in this test okay so what about power draw you ask well you tell me overclocked thread Ripper pushes total system draw to 414 watts and that's pretty extreme that said it does pale in comparison to the 7900 X at 475 watts then we have the 79 60 X pushing total system draw to over 500 watts and finally the 79 80 XE at 530 watts remember this is just CPU load we're measuring here so if you have a few high-end GPUs in your workstation you might want to go and acquire one of those zpm modules to power it due to the limited time we had to test these new CPUs I haven't really explored the cooling options all that much I simply to remember on our custom cooled x-29 test bed and got benchmarking so I'm yet to see how badly they punished air coolers and all-in-one liquid coolers at the stock clock speeds the 79 80 X the only push temps as high as 65 degrees which is certainly getting up there given that we do have a massive 360 millimeter right attached to the loop then once overclocked to 4.1 gigahertz on all course temperatures did hit 77 degrees which is quite a few degrees higher than what we saw from the 7900 X at 4.7 gigahertz okay so as usual we have quite a bit of data to make sense of of course it's not just about delivering the best performance I think rather most of you we more interest in how much bang for your buck you're really getting so that being the case let's take a look at a few price versus performance scatter plots first that we have blender and this is a lower is better scenario which is why the numbers at the bottom of the chart work backwards the faster a CPU the further to the right it will be and the lower the price well the lower it will be situated so ideally you want to be as far right as possible and as low as possible on the chart it's a little confusing so here is our line indicating the very best value CPU starting from the Rison seven seventeen hundred and going through to the 1950 X anything along this line offers a good balance between price and performance and of course anything to the right of the line is exceptional as you can see though the only Intel CPU to touch this line was the core i7 7820 X while slightly faster than the 1950 X we see that in terms of value the 79 60 X and 79 80 X II a shocking value but hey who's surprised looking at handbrake we see at the Intel Core i7 7800 X and i7 78 20x look to be mixing it up with the AMD CPUs whipping out our linear line from the r7 1,700 to the 1950 X we see that the 7800 X and 7820 X actually present as better value options when compared to the AMD CPUs even the 79 60 X moves closer to the line in this test though ultimately isn't great in terms of value while the 7890 XE is again horrible finally let's take a look at the price versus performance ratio and testing with Premiere Pro CC here we again see that all of the AMD processors look to deliver a similar bang for your buck while the core i7 70 100 X and 78 20x are in the mix as well throwing up our linear line we see that the core i7 CPUs with the exception of the 77 40 X do very well the new 79 60 X and 79 80 XE though while they don't do very well Adil when it comes to value unfortunately those price to performance scatter plots really summarize things quite well in my opinion the new Intel Core on ion series is blistering fast in every test but ultimately didn't end up being that much faster than the thread rip in 1950 X given they cost anywhere from 70 to 100 percent more it's hard to justify the often just 10 percent increase in performance over clocked these skylake X parts aren't worlds faster either they're only around 15 or 20 percent faster than the overclocked to 1950 X while they did draw around 20 percent more power so this puts Intel's new 16 core and 18 core CPUs in a bit of a jam if you ask me now that we've seen how they do perform it looks like the conclusion I draw my thread Reaper 1950 X in 1920 X review was pretty spot-on I said it a guess it looks like consumers will need to spend at least $1,400 us to acquire the I 979 40 X to match AMD's 1950 X and I seriously doubt the 40% price premium will be worth it this certainly looks to be the case now that we've seen the 16 core model in action I also noted that the other big issue the X 299 platform faces other than the extreme price is its complete lack of ECC memory support whereas thread Ripper supports ECC memory the Scarlet X chips don't how that means anyone serious about their workstation won't even consider intel's high-end desktop platform finally i concluded by saying unless intel was willing to move on price which it now seems very evident they're not i can't see why anyone would invest in the x 299 platform and while that's my honest opinion and this really hasn't changed with the arrival of the 16 core and 18 core parts in fact it's probably just cemented that opinion unless I had money to burn and I suppose even then I would really struggle to imagine a scenario where I would spend $1000 u.s. more so a hundred percent more on the core I 979 80 XE just again at best maybe 20% more performance time really would have to be money and a lot of it that said though there are certainly businesses that could justify the investment the less time staffs around waiting for a task to complete the quicker they can get on with the job so offering the world's most powerful desktop / workstation cpu intel will be more appealing in those scenarios the rest of us though us in Suzy estaba Blee going to be hard to justify spending this kind of money on these CPUs well that's about all I have to say for now I think that summarizes things pretty well Intel's obviously very competitive when it comes to performance I mean they do now have the world's most powerful desktop CPU that is at least until I'm damaged that thread rip is really nothing more than a disabled epic CPU and released a 24 core model just to stick it to Intel they say that's not going to happen but come on guys as if they're not gonna now is that before the end of the year I'm Yahoo Steve see you next time