Gadgetory

hey everyone welcome back to another ask GN episode as always that leave your questions in the comment section below for next episode or if you're a patreon backer you can leave them in the page around discord we also have a second extra patron to ask GN video that'll go up separately from this and will be available only to patreon backers so if you're interested in that one go to patreon.com/scishow gamers Nexus to get it otherwise we've got the full-length normal episode here so 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 BRMS this board was used in our recent attempts 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 so prior to the questions this thing this is the CPU or AP that we got in from AMD from part of that boot kit that we reviewed and if anyone was curious I haven't sent it back yet obviously I got an email reminder from the warranty program was all automated that just said you need to send this back within 15 days otherwise you're not gonna be able to use the service again so I mean I'll send it back I have really no reason to keep it but I was curious about what would happen if anyone was curious that's what happens they basically say you can't use it again so it doesn't look like I would be charged for it or anything like that but anyway that's going back soon first real question is an interesting one that I've been talking with about with AIB partners about lately this is from David Watson who said hey Steve do you think we will end up seeing and Vidia competing with aftermarket cards more directly by releasing their own non founders edition cards with new custom coolant solutions and heat sinks with double or triple fan designs I think it's fascinating prospect and when I feel NVIDIA has had a lot of thought bent on and it's really bound to you as much sooner rather than later because if they can get more market share they surely will and I honestly feel it's coming from Nvidia and could be massive for them so do you feel this is on the horizon I I don't this like I said there's something I've spoken with a lot of people about because you look at what what anybody has done and they've introduced reference cards which both vendors have had for a long time but they have introduced reference cards that are separately branded and also more importantly sold directly by Nvidia and he has done things kind of like this but not quite the same so the best example with AMD would probably be the fury X where there were no partner designs allowed it was only Andy's design that was it but you still ultimately bought the fury action say fire or Asus or whomever so this is a little bit different although not too much but they're there but they're different enough so Nvidia they have their founders Edition cards that they sell directly currently at a pretty good price considering the market because they've stuck to about MSRP and it's the only other way to get on mice RP cards is basically go to EVGA x' website or any of the other manufacturers who have a store and buy it from them because the retailers aren't gonna help you out so do I think it's on the horizon I don't know I don't feel like this is something we'll see anytime soon that said certainly on videos in a good place where if they wanted to make a better cooler they can do it it's it's not they make silicon products it's not that much harder to make something with two fans on it so it's not like there's a technical hurdle there the only thing that's stopping on video from making better coolers on the cards is the fact that they have partners who can do that for them so why put the expense in there's extra marketing like work that EVGA and asus and all them can do for you they can handle the entire warranty process they can handle shipping and distribution so there's a lot of advantage to having partners handle all of these things because it's all logistics and SKU tracking and management of inventory and management of warranty and RMA and all this stuff and Nvidia spends their money on R&D and engineers for the most part and their own marketing obviously geforce and all that stuff is all marketed but there's an advantage certainly to having partners so it's not like they provide no service and their coolers are pretty good of course but the coolers are probably the weakest aspect of the entire argument because like I said it's not that hard for a company like Nvidia to make it cooler like the ones that their board partners make so yes I've spoken with some board partners about the idea and this is just kind of like casual conversation in the past as something that people are distantly maybe anxious about but I don't have any reason to believe that this would happen anytime soon that said Nvidia sometimes does almost feel Apple like in some of their marketing and in the way that the company carries itself I'm not saying that's a good or a bad thing but in terms of building a product and having a really clear distinction of like this is the product by our company like apples got the MacBook or whatever I don't even know the process but they have their product and that's it that's the end of the line it's the start in the end of it and they build this whole marketing hype around it and they control all the messaging on it so there are advantages to both approaches of having partners and not Apple if you take that analogy certainly has its own challenges like with price but also the windows products have a lot of their own challenges as well so yeah I don't know I don't think this is something that's on the horizon any time soon but I guess the only valuable information I can provide here is that this is something that we've all kind of talked about behind the scenes and have wondered about but there's no concrete evidence to suggest that Nvidia would compete with his Ford partners directly at this time in fact when the founders edition cards launched and they got a lot of flack for this but when they launched the founders edition cards were I think $50 more than than actual MSRP for the board partner cards and so they were actually distant that they're further away from competing with those partners because their cards were more expensive than the board partners in most cases so yeah no evidence to suggest that this would happen anytime soon but it's been thought about next question thank Buddha says quick question regarding the 200 FPS limit in newer questionmark games is this applied per monitor if you're asked if you're using multi monitor or is actual a game that's locked to 200 FPS total it's the game for the most part so doom does this and I think maybe destiny I think destiny does as well so with doom specifically I know what the cause of that is doom limits to 200 FPS because of a physics issue in fact the doom T mid they batted for both teams and video and AMT at different points in the development cycle earlier on it was Nvidia Nvidia brought the doom team out to a press event I think for the GTX 1080 and they showed the game and noted that it was capped at 200 FPS because of a physics bug that the team was working on resolving hopefully before the launch of the GTX 1080 obviously that never happened it never got fixed they came out later to some a.m. the event I want to say was the Polaris 1 the same bug existed there was no further conversation about it so the limit your scene typically if it's not because it's a console port that's been poorly done or something if there's actually just a kind of random high FPS limit it's often physics Skyrim and Elder Scrolls games and fallout or great examples of this as well there's a 60 FPS you can unlock the FPS and if you put it not powerful enough set of hardware and lower the settings enough and really blast your framerate you can do pretty crazy things with physics just just because the game bases the physics calculations on the frame rate so locking it to max 60 resolves that but once you start pushing like 400 frames per second you can really break the games so yeah that it's it's not it is not to do with the monitors unless it's a vsync thing if it's game engine bound meaning you've turned off vsync and you still have a cap then then it's probably because of a game engine limitation often the physics if they do it based on frames next question Michael Morgan says can you demonstrate the end-user benefit of HP MMS memory over gddr5 or g5x on GPUs please so we've got we've talked about this first let me refer you to a previous video of ours that was about how much HBM costs this was specifically a video talking about the cost to manufacture Vega products and the HP M was a large aspect of that so in that video we make the point of why Andy chose HP M how they restored more or less required to use it for that particular product because although it would have been cheaper with gddr5 or 5x even it it would not compete as well it wouldn't it wouldn't compete so HP m has got a lot of advantages to it one of them is power consumption and I have all the specific numbers in that video I don't actually have them in my head anymore we filmed that while ago but power consumptions much lower with HP m per gigabyte in an ideal world where HBM cost is something we don't have to think about you would use HP m on basically everything because the power consumption is lower so from strictly a standpoint of being energy efficient which helps elsewhere like in thermals or just I mean there's no reason to waste power in being energy efficient HP M is superior especially for the performance you gots a bit more complicated than that with AMD specifically the Vega products and the fury previously were already such high power consumption that requiring that much more for gddr5 is starting to become a problem in terms of your power requirements for the whole board all of it not just the GPU and then also further depending on the type of memory you're using with the GPU and the GPU architecture it's possible to become memory bound and this happens with it's actually we have to answer your question can you demonstrate we have demonstrated it with the AM D products again because they primarily are the ones that use HP m with a nd products we've shown that increasing the HBM clock overclocking HBM rather than just a core clock in fact in fact to you down clock the core overclocking HBM will generally typically up to a point of diminishing returns get you greater gains and performance than the core and that would be the beginnings of indicating that there is a memory limitation it's not all there but in the other test we've published and we did a lot when Vega launched you can more or less draw the conclusion that it typically becomes bound by memory before anything else so in this instance because h BM is advantaged although they have fewer stacks of HB m on Vega they they want with two stacks instead of four so you are still running a thinner bus than on Fiji with Fiji you had four stacks so I think it was a 4096 bit bus but it was HP m1 so fewer bits per pin so you end up with I think it's two stacks on Vega with HB m 2 so it's a thinner bus 2048 or something but it makes up for it in the frequency of the memory which is much higher it's like 945 instead of 500 all these numbers off the top of my head I think they're pretty accurate though so yeah the the answer though is HB m is basically a requirement on GPUs that become memory limited or GPUs that have serious power consumption issues already where you can't afford another 30 40 50 watts for GE or 5 especially at the capacities that HB m can do in the same space for the power that it takes and we do have a lot of great content on that so just search the channel for HB m or for Vega and find some of the stuff we posted when it launched I think we also did some HP I'm testing on the Titan V where we same thing overclock HP M versus core what happens so you can find that content as well next question is Vishal Baabda who says why do CPU what why do CPUs not have different manufacturers like GPUs there are more manufacturers we might get more enthusiast features from the factory like liquid metal Tim and better IHS so first of all ideally you use solder not liquid metal liquid metal coming out of the factory doesn't make a lot of sense if they're going that approach it makes more sense for them to just solder it at that point so liquid metals more of an enthusiast thing as you've pointed out but it doesn't make a lot of sense manufacturing it would once they've gone that far solder it's where you go it makes more sense for the manufacturer better IHS sure that's that's certainly possible but yeah so the general answer here why don't CPUs have different manufacturers like GPUs consider the market I guess so with with GPUs you've got nvidia and AMD they make everything with CPS obviously Intel and AMD make everything but GPUs require a whole lot more than just the the single part you don't just buy a GPU from Nvidia or from AMD and then socket it into a board like you buy a CPU and you socket it into a motherboard so a GPU is is fitted with enough special requirements that it makes sense to have those board partners you have different sources who supply the memory you could get it from at one point al-qaeda or these days Samsung hynix micron for example so someone needs to take control of where the memory is coming from make sense to put a manufacturer there they pick the memory supplier they pick a PCB supplier they design the PCB they design the cooler they pick suppliers for the cooling components so there are a lot of suppliers that go into a single GPU or a single video card I should say and I suppose you can make the argument that why don't companies like Intel just make the motherboard because that would be the equivalent it would be if they if Nvidia and AMD made their own cards then Intel would make their own motherboard well an Intel did make their own motherboards at one point it didn't go too well they they had a lot of issues controlling it and part of that I think does come down to what I was saying earlier with Logistics and part tracking warranty all of the supplier relations certainly a company with a market cap of 150 billion dollars or whatever it is these days could make that stuff work but they tried and they didn't do a great job of it that until motherboards went under they even had that whole better together branding as well which AMD stride similar types of branding but I I don't have a great answer specifically why but from industry experience if you kind of look at the whole picture of things it makes sense to me that with something that's a socketable part like a cpu you just make it you sell it direct cuz cut out the middleman kind of like what we were talking about with Nvidia why doesn't that video cut out the middleman well there a couple of reasons to keep them around Intel tried to cut them out with their own motherboards didn't go over so I don't know I that's kind of how I look at it Intel also if you kind of look at how they're making the parts and many an AMD don't fabricate their own GPUs they go through partners Global Foundries tsmc people like that Intel owns their own fabs which is quite a bit different because that's all in-house and there's I mean they already don't have middle people to make the thing so I have middle people to brand it and put a heatsink on it I guess because that's really the only difference you have at least with GPUs you have a couple more things you can specialize so yeah I don't know that's that's how I see that guess next one we have a celebrity question this is from Satoshi Nakamoto I believe it's the real one we've we found him he was on our discord I don't think that surprises anybody Satoshi says of course hey guys could you reach out to Thermaltake and ask them if they have any idea for the arrival of the level 20 case well mr. Nakamoto if I may if I may call you that the level 20 case I actually I sent this question the Thermaltake not sure whether it was a joke or not and apparently it's in high demand a lot of you want to know it's gonna be available there's not a hard date yet they're pushing to get it out keep your eyes on their social media channels they're just finalizing all the production and everything like that so shouldn't be too long for anyone who who doesn't know that's a $1,000 case so it would make sense that the god of Bitcoin I guess would want one next question defenestration eyes throwing people out of windows says Steve a massive limit for AP use is there need to use system memory do you think Abe use will remain on the low end or end up high end is there not actually limit to ddr4 ap use more memory channels on AP use or will more memory channels appear in the future on AP use given how board partners operate and push for chip consolidation do you think we might see a motherboard RAM GPU CPU as a single PCB I think there's a couple questions here so first one do I think ap use will end up in the high end how do we define high end if our definition of high end is competing in graphics capabilities with a 1080 Ti than know if our definition of high end is competing with what I would call mid-range maybe let's say a 10 60 or an RX 580 or Vega 56 something kinda in the 200 $350 in an ideal world price range I still don't really think so not today not today but ap use are cramming a lot of processing power into a small space and he's done a great job where they realized that these are for gaming devices into outputs in IGP and everything they make basically they take less die space to put that IGP in there then Andy does for their graphics components because Intel is probably under the assumption that you're buying the i7s and coupling it with a d GPU Andy took the other approach of saying if you're buying something that's 300 plus dollars then screw it we won't include an integrated graphics component we'll use all the space for the the CPU and then you can go buy your own graphics card and deal with it yourself and then they also make the assumption that at the low end with ap use you probably if you're spending one hundred two hundred seventy dollars on it are buying it for gaming and so if you're buying it for gaming the GPU component sort of matters the most provided that you have enough frequency and a couple of threads from the CPU so with that provision I don't think that Andy is going to be cramming like a Vega 56 equivalent into an APU anytime soon look at how much space Vega takes 56 takes right the die size alone is a concern you start eating more of the shared die space or the shared substrate for a bigger graphics component at some point you run into limitations as we advanced in the future it'll get that direction I don't I don't think there will ever be though in the immediate like next couple years I don't think there will be an APU that would go properly high-end like 1080 plus in modern day terms because although an APU and maybe five years could do what a 1080 does today I don't think it's gonna do what an AMD like Vega 76 or whatever does at that point in the future next question though but they are good though the APS are good they do the the things they're supposed to do but it's not meant to be that kind of component it's you're asked you're asking to cram too much into one spot I think could happen in the future though distantly next part of your question was let's see consolidation a single motherboard Ram GPU CPU PCB it kind of exists to some extent you can get basically PGA devices you look at laptops that's more or less what you're talking about take the Apple example again Apple solder is it all on to one board for the most part soldered CPU it's all BGA ball grid array soldered memory soldered flash components in some cases it's all soldered so I that's I guess kind of what you're talking about at that point it does exist typically in really small form-factor applications we're having modular components doesn't make sense because of the space requirement or the thermal concern see next one so this one I need to we'll come back to this more in the future once I get around to the fans housing content at some point we do the basics now Thane of Starfall said I've bit confused by case fans lately specifically rpm and in relation to voltage if I understand correctly different fans have different max and minimum rpm at a given voltage but what happens when you connect two fans with different rpms to a single header on the motherboard using a Wye splitter so yes you do understand correctly if I take 10 Noctua nfa fourteens which we've actually done and plug them into the same motherboard header and spin them up to max 100% versus whatever the minimum voltage is you'll see different rpms and that's manufacturing tolerances as for I think part of this question was about fire I've read on forums that the cable or port can catch fire if they draw too much power is this true it's always possible make a fire with electronics but it shouldn't really happen so what you're asking basically the the fan headers on motherboards are typically rated for one amp and the high amperage or H am P is typically what it's labeled high amperage headers on motherboards are often rated for two amps not always check your motherboard manual but if you put a Y splitter into a 2 amp header you're probably gonna be fine and you plug litter into a 1 amp header and then you plug a bunch of fans into it and run them at max speeds you might have problems theoretically shouldn't start a fire but but yeah it comes on to the to how much current the header can handle it's one or two amps typically so just be reasonable with how many fans you're plugging into the Y splitter and then you had some questions about PWM stuff we'll come back to that with our fan content in the future go more quick ones one of them so this one how do dealing with from ash borer how to delete it with liquid metal temperatures compared to soldered CPU temperatures do you ever plan to deal it a risin CPU and test results under the assumption that D lighting provides better temperatures so I don't mind that Intel doesn't solder anymore as an enthusiast I want to do it anyway and if it's not soldered it's easier to deal it if it's if it's soldered you won't have to deal it is more or less the answer to that question if it's soldered they've done all the work for you there's no reason to have liquid metal it's gonna perform within a couple degrees in the best-case scenario and solder may be a bit ahead and maybe a bit I would assume a bit ahead dare Bower did a video on this he deal it at arisin seven CPU and the variance was not that great so yeah I mean it same what I just said is all that needs to be said I guess if it's soldered deal it in is basically irrelevant at that point next question armand be mod my out of stock damn miners yes well the gamers Nexus anti-static mod mat is well known for being used in factories in China that are filled with GPU mining equipment I guess if you want one though they're on backorder and they will be shipping end of this month or early next month so get them before the miners do I guess last question nori the explorer.exe what is an important fact viewers should know about gamers Nexus and the opposite what is a big misconception viewers have expressed about gamers Nexus important fact if I'm taking it seriously is probably that we've been doing this for about 10 years so a lot of experience in the game started with gaming content almost exclusively didn't really get into PC hardware stuff till maybe 2010 in terms of like editorial work I'd always worked on it personally but PC hardware maybe 2010 2012 areas where I started getting more serious about it and basically pivoted the site to hardware coverage instead of games coverage gaming coverage there was understand I have a lot of contacts in the game industry still but we always don't talk to them as much anymore it was interesting because it gives you a really good understanding of things like game graphics game programming game engine limitations all these things that if I had started in state and hardware I wouldn't really understand and so knowing a bit extra obviously there's an impossible now to know but knowing a little bit more about game development and game engine development from those starting years before we got into hardware has been really helpful in understanding how to test different aspects of games and in understanding how the games interact with the computer hardware what happens when you toggle different settings on what do the settings names even mean that's all stuff I learned from the early days covering basically games exclusively about ten years ago so I would say that's an important fact that people know we started as a Games coverage outlet and it has provided me with a unique perspective on Hardware coverage that I think we've been able to leverage with some things like our Final Fantasy coverage recently misconception I think Budokan in the ask GN basically answered this for me something about shredding are we I'll use that as a as an opportunity to plug the GN Steve side-channel where we've been uploading our patreon behind-the-scenes videos some of them go live subscribe to that channel if you haven't because every now and then we'll set a patreon behind-the-scenes video public for everyone and that's where it will exist so if you're subscribed there you'll get it as far as shredding our I've got some some downhill mountain biking videos over there where I've done things like talk about coffee like and rise in the while riding down the mountain at Whistler and I don't know we have some a line footage or something we can show while we're at it so that's it for this one is always subscribe for more go to patreon.com/scishow and resent axis dogs out directly or get the extra version of this ask GN and stored on cameras axis dotnet to pick up one of our mod mats on backorder they'll be here very soon for the second round so thanks for watching I'll see you all next time