Gadgetory

CP is with integrated graphics always make memory interesting memories commoditization ignoring recent price trends has made it an item where you sort of pick what's cheap and just buy it with something like a and these Raven Ridge AP use that memory choice could have a lot more impact than in a budget gaming PC with a discrete GPU we'll be testing a handful of memory kits with the r5 2400 G in today's content including single versus dual channel testing where all timings have been properly equalized we're also testing a few different motherboards with the same kits of memory useful for determining how timings change between boards and we'll be getting through gaming benchmarks primarily today rather than anything with synthetics before that this video is brought to you by EVGA and the X 299 dark motherboard for the Intel high-end desktop CPUs the X 299 dark is one of the only motherboards on the market with proper vrm cooling we've tested this and found a significant performance increase over those without active cooling on the prm's this board was used in our recent attempt to set a top-10 record in fire strike and you can learn more about the x-29 dark at the link in the description below the only place we're making use of synthetics will be between switching boards for the most part with fire strike and times by this is where synthetics shine we're avoiding things like Cinebench or non synthetic production workloads like blender because they just don't really fit the content we're going for how does memory impact gaming performance with the Raven Ridge AP use so to that end it's not like we're testing every metric on the market but we're going through a lot of them and covering everything from 2400 up to 3200 megahertz and we'll just kind of see how it looks and then decide what we're doing for future content pieces from there we're splitting these benchmarks into two sections first we'll show the impact of various memory kits on performance when tested on a gigabyte gaming k5 motherboard and then we'll move over to demonstrate how a few popular motherboards affect results when left to auto XMP timings we were focusing on memory scalability performance today with a baseline provided by the previously tested G 45 60 and r3 1200 with GT 1030 that we ran a few weeks ago we'll get to APU overclocking and a future content piece but for now we're just looking at memory before ins for single-channel testing we're benchmarking the best kit that we have out of all the tests we ran which is the Trident z CL 1430 200 megahertz option and then we'll look at a couple of the others along the way rocket league at 1080p and high-ish settings will start us out for this one our r5 2400 G operated a 58 FPS average with 3200 mega Hertz CL 14 Triton Z memory from g.skill and the 0.1% load is tagged LAN at 18 FPS the GT 1030 and G 4560 would be a good baseline here for comparison performing at about 62 FPS average but was significantly improved frame times remember the memory for an APU is shared between the GPU and CPU components when the GPU component on an APU needs to access memory it's going to system memory there's no onboard memory like you have with gddr5 or HB m2 and because of that we're dealing with different types of memory there is a significant difference between gddr5 which is physically on the board it has tighter tie means typically and it's technically a bit faster for the specific task also it's physically closer to the GPU than an APU where it has to travel across part of the motherboard basically so this is where the differences start to emerge with memory and it's why memory is so important with APU is particularly when you're talking do-overs a single channel this is something that as we've shown recently doesn't really mean a whole lot when you use a D GPU and a low-end CPU you could run one stick or two sticks and as we've shown in a lot of cases it really doesn't matter that much however for something like an APU it can matter a whole lot as we've just shown here and that's because the APU actually really needs access to memory because the GPU otherwise would be starved for it it's accessing system memory its shared memory between the CPU and the GPU and so it needs to be fast enough to support both of those devices simultaneously back to the rocket league chart keep in mind that our focus today again is memory so we'll get to APU overclocking in a later content piece getting down to the memory comparison on the r5 apu the next best performing kit after the CL 14 Trident II was our corsair lpx memory that AMD provided with the r7 last year operating at 29 30 megahertz NCL 16 the next one down was our guile kit also provided by AMD before the r5 launched operating at 3200 megahertz and CL 16 in this instance however there's more different than just the top-level cast timings we'll get to those toward the end of the content as it's decidedly more boring than FPS numbers for most scalability has this ranking the LP axe and guile kits within margin of error of one another effectively equal in performance the 24 hundred megahertz CL 15 Ripjaws 5 memory from years ago operated at about 48 fps average marking the top to bottom range of performance as about 9 FPS between the Tritons ii and the Ripjaws kits that plants the 32 hundred megahertz CL 14 tried and zeke hit as 19.8% ahead of the Ripjaws kit and the Ripjaws unit also dropped frame times further landing at 13 FPS 0.1% lows finally our single channel test takes the best kit the trident z CL 14 memory and tests it with only one channel firing this lands us at 31 FPS placing the dual channel tried and zeke it at about eighty percent ahead in performance we nearly double performance by operating in two channels or in most cases adding a stick of memory if you're considering buying one versus two so far with only one data point it looked like builders of apu systems will definitely want to go dual channel and dual stick even if it's a couple dollars more the difference tends to be 5 to 10 bucks so it's worth it here for dota 2 we test at high settings and 1080p using DirectX 11 note that and the rise in CPUs have traditionally fallen behind in this particular game so we would anticipate that performance behavior would follow through on Raven Ridge this is more a factor of dota 2 and how it's built as it favors frequency over threads in a lot of cases regardless our 2 GT 1030 tests were bound by the GPU topping out at 63 FPS average the our 520 400g with triton z memory maintain an average of 45 FPS with it lows somewhat low at 19 FPS the 29 33 megahertz LP x kit falls in second place again for the second time with the Ripjaws 5 kit outpacing our guile see all 16 memory by a few FPS the stack places and Alpha X memory has functionally tied and within error margins with the guile kit not far behind going single channel nukes performance pretty an output down to 25 FPS from 45 FPS granting the dual channel configuration a 79 percent lead over the single channel configuration those aren't much different but that's largely a factor of how this game generally behaves for csgo with relatively high settings at 1080p the baseline tests operated around 111 to 120 FPS average with our highest performing our 520 400 G non over clocked at 95 FPS with the Tritons II memory note also that the r5 lows maintain a much healthier number this time marked at 41 FPS for 0.1% and 63 FPS for 1% lows csgo is fully playable on the r5 2400 G and maintains a consistent frame to frame interval during gameplay the next component was once again the courts are lpx at 29 33 megahertz with CL 16 timings allowing the Triton Z to lead by about 5.9 percent the guile kits rank roughly within error of the Ripjaws 20 400 megahertz kit further demonstrating that this particular kit of memory doesn't play as well with this specific motherboard keep in mind that explicit motherboard tuning for validated memory kits can impact performance so other boards may behave differently it's entirely possible that one board would not like the Triton xik it at all but love the guile kit and vice versa the Ripjaws 520 400 megahertz kit places lasts about 10 FPS behind the Triton z kit which results in a lead of about 12% for the Trident II memory pulling up Newegg the price difference is about $230 with a trident ii kit at 16 gigabytes or about 165 for the Ripjaws 5 kit that we used previously again that's at 16 gigabytes and strictly committing the exact two kits that we own back to the chart single channel testing how to us at 52 FPS average 0.1% low is significantly reduced we have more frame time variants here than previously we're about 83% ahead with the dual channel alternative Triton C kit Sniper Elite 4 gives us a DirectX 12 test with a sync compute and one which has been exceptionally well optimized especially for AMD we're testing at 1080p high as that's what we have the most data for with standard GPU benchmarks note that of course you'd probably dial down to medium for actual gameplay but we're just looking at scaling today for this benchmark we observed the 2400 G outperforming the GT 1030 and G 4560 combination notably which comes down to an AMD advantage with sniper elites asynchronous compute pipeline particularly on the Vega graphics chip the Tritons ECL 14 kit ranks about 30 FPS average with it lows tightly behind and the 29 33 megahertz kit CL 16 nearly tied in performance we're beginning to bump against other limitations here the CL 14 tried and xik it again lands 12 to 13% ahead of the 24 hundred megahertz Ripjaws 5 kid and the guy'll kit again demonstrates poor performance from its auto timings on this board single channel performance is significantly behind as well it's just like previously and just briefly here's an overwatch chart as well we'll talk about this more in the article below but there's not much to say here it's more of the same scaling that we've seen previously when tested at medium settings with 1080p firestrike gives us a synthetic recap to look at everything including a quick test we did with a 16 50 megahertz overclock on the AP is Vega 11 GPU which roughly equals or slightly under performs against the GT 1030 as for memory performs the Trident Z kaitlin's our total score about 32 89 which is about 3% ahead of the Corsair lpx kid or about 5.8 percent ahead of the guile kit and about 11.6 percent ahead of the 24 hundred megahertz ripped off 5 kit versus a single channel we see a market uplift of 59% in total score and depending on whether you're looking at GP or CPU that may increase or decrease for time spy will just briefly mention that the overclocked Vega 11 component does well to place ahead of the GT 1030 build but we'll talk about that more in a later piece focusing back on memory the Trident II kit places our total score at 12:55 which ranks about 11% ahead of the Kyle 32 hundred megahertz kit that AMD provided for the RFI review or about 3% ahead of the 29:33 kit once again single versus dual channel configurations have a tremendous impact on performance as you can see here movie from this the next test is pretty simple really we didn't go crazy and depth on it you could really go off the deep end here and test every motherboard on the market if you wanted to all we did was take three or four popular motherboards that we happen to have and run the same two kits dial and try and Z through fire strike and times by with an r5 2400 J the entire point of this is to determine if the boards behave any differently if they have different auto timings applied when you just apply XMP because it doesn't always apply the same settings every single time and we have a separate table for that after we get through these last two charts to show though maybe one or two instances where those numbers are actually different aside from just saying the word Auto four times by the gigabyte gaming k5 with triton z memory ranked again at 12:55 for its combined score with the pro four from asrock within our margin of error these two are effectively equal and the msi tomahawk scored at 1234 the tried and xik hit giving us a top-to-bottom scoring of two percent difference fire strike positions the gigabyte gaming k5 at the top but also completely within error of the asrock pro four we were within about 1% of performance and so these boards apply timings in a way that makes them perform functionally equivalently the tomahawk p3 50 board lands the Trident Z kit also close behind between these three boards there's nothing crazy going on in score they're all about the same there's a slight difference in timings though here's a simplified spreadsheet on the motherboard auto timings for the Triton Z kid note that of course different memory kits may have very different results here the Triton Zeke it is fairly popular and so most boards have explicit support of this memory our main difference is in TR FC he has shown here though that doesn't manifest in huge ways in these quick synthetics as for memory timing settings on the main gaming k5 mother would publish those in the article link below so the main takeaway here is well one single versus dual channel does actually matter a lot with ap use that's not really news and it's not quite so important for something like a low end CPU with a D GPU we did that test a few months ago and for that configuration didn't really matter but for this one it does and that's because of how ap is worked because there's no onboard memory with it so if you are bought it doesn't matter what kind of memory you're buying you're planning to play games in APA which you probably are so they're targeted at then we would suggest find the kid of memory you want and then find the version that has two sticks so you may be two by four if you're going with the net gigabyte stick by two four gigabyte sticks instead go dual channel it will help it will matter and often times it's five to ten dollars more for two by four instead of one by eight if that might even be a couple bucks cheaper it depends on like kiddo memory as for the frequency it's obviously not quite as simple as just frequency timings matter the most and more than just the couple like four or five top level timings sub timings and AMD boards can get kind of screwy depending on what board you're using so there's a lot of play there we can't test for every scenario but the g.skill tried and xik it did really well in our testing it's also pretty expensive in fact it's it's kind of exiting the price range of what's reasonable to buy with an APU if you're spending that much on memory you almost might as well just get a d GPU and a cheap CPU so you might have to dial it back a bit the good news is even coming down from that expensive kit of memory there's not a whole lot of range in the following sort of one to two kits that we tested the Corsair 29:33 kit that we got a year ago from AMD for Rison because it was at the time the only valid aid kit Verizon that kit did pretty well here it's often roughly tied with a tried NZ kit it's reasonably affordable and g.skill has good Ripjaws kits that are faster than the old 2401 we tested here you pick one of those up the point is until you really get down to like 20 426 66 megahertz you're not losing a ton of performance so if the difference is significant in price which it can be especially for countries that aren't the US where prices we've looked at them the memory prices are kind of crazy there too but if the price difference is a significant between I don't know 29 33 on 3200 just get 29 30 it's fine you can apply that money better towards something else because the minute you're spending an extra twenty dollars on memory you're really getting into territory where a low-end CPU and addy GPU with lower end memory would be better value for your money but if you are somewhere I mean the u.s. sometimes has about the same price between thirty two hundred and twenty sixty six if you're in that situation I don't it's I guess just buy the better one because it does actually matter a bit so yeah not not world ending differences in the two primary use cases that we tested the LP accent right and Z but pretty significant differences between the Triton Z and the Ripjaws twenty four hundred megahertz kit so that's where it starts to come into play and those timings were pretty close so you are starting to look at a raw frequency difference to some extent there and again this isn't the most comprehensive in the world that you could do because there's a lot of memory there a lot of motherboards they all kind of play a little bit differently through your own research all that stuff but hopefully this gives you a starting point and some basic concepts as to what really matters with the Raven Ridge ap use and as always you can subscribe for more coverage of these AP use or you gonna store it on Karen's Nexus dotnet if you'd like to help us out directly by buying one of our restocked mugs I think we have hats coming in too and patreon.com slash gamers Nexus to help us out other ways thank you for watching subscribe for more I'll see you all next time