Gadgetory

x3 seventy verses x4 70 chipset differences are pretty simple there really aren't that many at least not that many really important ones are critical and you can still use Rison 1 & 2 on each platform if you want to it just comes down to the vendor to properly support the CPU that you're using make sure there's good memory support for it and that's really the crux of the differences its memory support so in this test what we're showing as X 370 versus X 470 but the air quotes are there because ultimately it comes down to all the million BIOS variables and how well each motherboard vendor has tuned for them before that this video is brought to you by thermal takes view 37 case the vo 37 focuses on highlighting custom PC builds with its full panoramic window and tinted front acrylic and our thermal testing the view 37 performed reasonably well when considering its looks focused build which is partly thanks to the airflow design and the removal of a bottom power supply shroud for a balance of looks and performance check the link in the description below for the view 37 unlike Intel chipsets where basically every generation of chipset has some kind of semi major change maybe the high-speed i/o Lane count changing from 26 to 30 and these are pretty straightforward we have a chart of the i/o differences quick tip there are none and most the rest of the differences really again it comes down to memory support and overclocking features not necessarily memory overclocking as we showed in our review you kind of get stuck at 36 66 maybe you'd certainly push it higher under the right conditions but for most users that's where you'll get stuck maybe one tick higher so memory overclocking itself hasn't really improved a lot it's specifically memory support for different kits and we will demonstrate that today with guile versus Corsair memory where we with one kit saw very large differences in X 370 versus X for Sony frame rate performance and the other kit we saw basically zero differences so what does that tell you that tells you that the chipset itself is not that relevant to the equation of performance because when we talk about motherboard differences or chipset differences unless you have the exact same motherboard in both the last generation and the current generation exact same 100% identical 110 identical BIOS everything unless you have that and then you just change the chipset you can't do a proper a B with the chipsets so consider this more of a how have the motherboards changed as well have they improved at all like Andy and the vendors would tell you they have with memory support the only other difference here is going to be precision boost and store mi outside of the BIOS changes and all that stuff in terms of layout memory traces have changed a bit and the rest is basically just coming down to updates to UEFI and memory support our previous video the 2700 X review which you should watch showed export sony memory testing for frequency differences where we found that there was about a 1400 point performance range and times by when controlling for timings that range dwindled down to 200 points which we can also show when we left the sub timings and tertiary timings auto and adjusted only the frequency and controlled for the primary timings the point is that these boards are getting sophisticated now they're getting good enough to change the subtitle the tertiary times that most people ignore don't pay attention you don't know what they do so that they can accommodate for less strain if you have for example 3600 K that you drop to 3200 it might actually tighten up the timings for you so you have to control all those variables but that was where we started with this let's go into the new data for our final data in the review we switched over to a kid of Corsair Vengeance 16 18 18 36 memory at 3200 megahertz we performed initial testing which was not published on our guile kit that AMD previously provided but ultimately discarded the data as the guile memory performed significantly worse on both Intel and AMD platforms but mostly AMD we have two sets of tests to look at guile memory testing with Auto sub timings and manual primary timings and coarser memory testing with controlled timings the motherboards used are the gigabyte X 470 gaming 7 and the AES crosshair 6 hero X 370 motherboard for watchdogs - starting with the 1700 X we observed 2-0 difference when using controlled timings and Corsairs vengeance lpx memory which will highlight with all differences landing in-between error margins these are functionally the same you'll note that the error bar is on the charts are aligned with differences measured meaning that we can't meaningfully determine an advantage 4x4 70 or X 370 using guile memory the kit that Andy provided with the r5 1600 X last year we observed about a 10 FPS change in performance between the X 370 and ax 470 this is primarily a result of sub timings and tertiary timings which some AMD boards still have a lot of problems with it comes down to the board vendors make sure your kits are validated for the risin motherboard that you're buying and you should be fine those 10 FPS differences though manifest as a 10% swing and performance 2x3 7e favoring its but all this really means is that e crosshair 6 hero has better tuned timings than the gigabyte board we validated this by looking at the configurations and noted some differences in user controlled auto timings there are also timings that never get presented to the end-user so we can't be sure how those are configured for the 2700 X with guy'll memory we also observed about a 10 FPS performance wane from 89 on guile with x4 70 to 100 FPS average with guile on x3 70 again this isn't the chipset it's the motherboard and that means every board is going to have slightly different behavioral characteristics we ended up testing only with coarser memory as we found it to be best tuned on all boards without requiring as much extra technician oversight for total war Warhammer we observed the Corsair kit producing no significant difference between the two platforms controlling for timings functionally produces equivalent performance when stock the guile kit demonstrated 83% performance difference which is small even for that kit the 2700 X produced similarly nondifferent results with the guy'll kit we noticed about a 1 to 2 FPS delta which is with invariance civilization 6 is tested for turn time instead of FPS as the FPS increases with slower CPU is due to lengthy turns sitting static FPS is also a useless metric for turn-based strategy games return time we observed functionally zero difference between actually 17 X 4 sounding for the 1700 X demonstrated here we measured a 0.1 second difference that's within variance for the 2700 X we also measured node we are within error margins even with the guile kit for project scars all kits on the 1700 exit landed at around the same performance this game is not memory intensive which helps further illustrate our point that memory sub timings are primarily responsible for performance not the chipset if you tuned every single setting in BIOS and use 2 otherwise identical motherboards you'd get about the same performance the same is true for the r7 2700 X with guile memory here where we observed performance differences within a margin of error ashes of the singularity tends to be somewhat memory intensive and shows the same exact trend on the 1700 X as we saw previously there's no change from other games for the most part the 2700 accent ashes is also similar where we've seen better performance on X 370 with guile as a result of better memory auto timings on a Seuss's half as demonstrated in the previous chart with the coarser memory it all equalizes once you have controlled timings blender produced the same kind of result scaling he saw his 0.1 minute difference on the 2700 X when tested with Corsair memory timings controlled there's no chipset impact at all he saw a 0.2 minute difference within error with the 1700 X so there you have it differences emerge from things like the board quality primarily not just that we could not find a simple chipset difference in terms of your frame rate or things like that not with a 1,700 ax not with a 2,700 X and yes it's entirely possible there are differences that are more worthy of note but we could not find any for performance it's not we obviously didn't test every single application in the world that said for consumer gaming things like that don't sweat it the thing you should be concerned about is the vrm quality for example the gaming sevens got a pretty good VRM but it's also kind of weak in the BIOS Department they're getting better they're good enough for most users but both of those parts of the equation matter a lot and that's really what the difference comes down to is who's got the best memory support for the kit you're buying and you should check every vendors memory support official validated list when you buy your memory that's something that I think a lot of people probably overlooked in the building process because it seems like memory is kind of memory but that's I mean although that's true to an extent the motherboard vendors are ultimately responsible for making sure the memory you're using is properly configured for the lower level Auto timings unless you want to do that yourself and that's the takeaway it's for the most part there's no difference a to be with the chipsets that we can presently find and to properly test that you would need basically this motherboard in identical form and I would need a socketable chipset that I can just pop in and out as I desired for testing that's what you'd really need for a perfect test but we got close enough to tell you that the motherboards are the difference not really the chipset which is more or less what we said previously and our pre review content and a review content so that's it for this one subscribe for more as always go to patreon.com/scishow and axis top that directly go to store dock gamers Nexus dotnet to pick up a mod mat like this one we're really happy with how their sign doing quite well we'll have another round of them in the next couple of weeks shipping out to you guys so thanks for the back orders and for watching I'll see you all next time