Gadgetory

hey everyone welcome back to another ask G episode if you have questions for next time leave them in the comment section below and please try and upload each other's actual questions so that I can find them rather than waiting through a bunch of memes about cats which makes it really difficult to find the actual questions so this time we have a couple good ones on by or wait for Rison + / risin - a common one did you make ROI on the Titan V we also had a question about the GN mod mat and it's spill resistant properties which required us to film an infomercial first we've ever done we hope you enjoy before getting to that this content is brought to you by the Thermaltake flow RGB closed-loop liquid cooler which is a 360 millimetre radiator + 3 120 fans that are RGB illuminated if then we'll take it ringg fans at that this is a 4.5 done a stack pump which one of the faster pumps you can learn more at the link in the description below so let's do buyer weight first again we did this for the 4 Volta last time I think largely the same answer here basically if you need a computer today then you should buy stuff today the prices aren't gonna change that much at this point it's looking like the rise in sale prices are more or less becoming the rise in prices the retailers have only increased them once or twice the last couple months back to where they were and rising seven seventeen hundred and five sixteen hundred have state basically on sale almost for the last two months now so I wouldn't worry too much about the prices going up they might a bit after Christmas and the holidays but either way these are still good processors and there's no reason to feel bad about buying a good processor if it does what you want it to do obviously like any production tasks or blender or video rendering or any of that type of stuff that might take advantage of those cores they're good for that already and Rison plus will be a refresh in the same way that some of Intel's plus to market demarcate lines have been refresh is an Intel sort of started that with their plus nomenclature that they just stick at the end of an existing fab process to make it sound like it's more advanced than it really is and yes they changed things like like fin height and gate pitch and things like that but ultimately it's not a massive change so and it's gonna be the same Verizon Plus that's it is advancements on an existing CPU that means it'll be things like tuning frequency perhaps or tuning memory timings or memory compatibility in general or things of this nature that have been done or learned about over the past however many months since it launched I guess about nine months at this point so yeah the answer to wait for a rising plus or buy comes down to do you need a computer today and if you do then buy something the way I always do this and I think I said this in the last one too is I would suggest asking yourself first am I happy with my current machine the answer to that should be contingent on does your current machine do what you need it to do in a reasonable amount of time because you're asking about rising plus in this scenario I would assume it's probably for some kind of production machine not just gaming so if that's the case you should be looking at how long does it take me to render the things I need to render and then go find some benchmarks how much faster is rising now today is that speed benefit something that will benefit me now today or can I live with what I have that's really all it comes down to now timelines Rison plus and rising two is really far away there was a headline in our news video last week we had a source what's a couple article sources that said rise into first quarter 2018 that's objectively incorrect it's really easy to prove you can look at am these own timelines the ones that they gave us in March at the Rison announcement and launched and it was not rising - in first quarter of 2018 I can tell you that right now so that's wrong what we talked about in the video the news video I basically took that and without really explaining why I called it Rison plus that the reason why risin + will be coming out sometime in the first half now I don't know exactly when but I know that they're gonna be talks about risin plus at CES we're not under embargo for any of that in fact I don't think we've currently scheduled a meeting but we'll work on one but yet there will be conversations about it soon within the next few weeks I don't know when it launches but the answer is still the same though it's you know you're looking at a launch after probably January so can you live without a computer for a month if so wait if not by now there's no reason to feel buyer's remorse about buying a processor that's still perfectly good and probably a bit cheaper than what's going to come out with Rison plus so that's how I look at it that's also why I think the prices are basically permanent at this point they're trying to leave room for when Rison plus comes out which is better I guess than what intolerant really does they just keep everything the same price but I don't know hopefully that gives you some idea of timelines and things like that you're looking at first half for sure probably so that's kind of my understanding right now and then the rest is personal preference did you make ROI on the Titan V this is a not asked by any particular person we had a couple different people asking this so the Titan V was a three thousand dollar business expense I say that specifically for a few reasons a business expense is not you just dropped three thousand dollars on a video card you can afford lots of things no I did not spend three thousand dollars on a video card my company spent $3,000 on a video card there's a big difference there one of us can afford it the other one can't so the company bought an expensive video card as a business expense the very short answer did you make ROI is yes the longer answer is and I'm only talking about this because I think a lot of people are interested in it the longer answer is there's different forms of ROI you can make as a media outlet and that would be things like direct ROI when you watch a video you might be served a YouTube ads add or you might be YouTube red subscriber either of those situations gives us direct ROI return on investment where you're generating probably fractions of a penny with your views on that content we can get a lot of views on that content that's a lot of fractions of pennies but it doesn't equal $3,000 so there's also indirect ROI that would be things like the easy one affiliate revenue this product doesn't generate a lot of affiliate revenue it's a $3,000 video card that doesn't fit our core audience so we kind of ignore that one there's where affiliate revenue would apply would be things like more affordable like cases and things like that there's also direct ad sales that would be when we place an ad like the one that was probably at the start of this video unless it was for one of our own products but we have ads for thermal grizzly Thermaltake we've run I fix it we've run EVGA we've run AMD in the past we've run Nvidia in the past so all these different advertisers are the direct sales that would still kind of fall into the bucket with Adsense ads where they go together but it boosts the purview basically dollar value it's still gonna be low you're still talking often well always pennies or fractions of pennies per view but then there's other forms of interactive ROI like we sold a good amount of ads for December because it's end of year this is pretty normal end of year everyone is all the businesses are spending all the rest of their budget for the year basically every business because you're trying to get stuff in before tax season or you're trying to get if you're a large company you're trying to spend your marketing budget or whatever it is so that you can claim you need it next year when they revile you ate your budget if they say you had ten thousand dollars left so we're gonna give you ten thousand dollars less this year so everyone's trying to spend money so we sell more ads and we need to place those on videos and make sure we get views on them so that's kind of indirect there's also finally it's this was large about a new architecture and learning about it and trying to get a head start on what we do when the next actual gaming architecture comes out we did this with Vega Frontier Edition as well it gives a huge leg up in terms of knowing what to expect and what quirks there will be in a new architecture so now we have all this experience with Volta and when it's not it probably won't be almost certainly what we call it Volta but when the gaming derivative comes out we'll have a bit of an advantage we already know all the quirks of Volta so if any reasonable amount of that finds its way into gaming we already know it's test that cuts out a huge amount of work for me in a couple months and it puts us a bit ahead of competition so that's another form of indirect ROI and then there's things like just you know we you become one of the only sources of information so people in the industry are watching those videos not just enthusiasts but you're looking at industry contacts watching the videos and that has potential for ROI later down the road as well because they see you as willing to invest in your company to prove a point or to learn about something and also sort of sends a message that hey if you don't send us the thing we'll buy it anyway so you can't really hide it from us so there's a lot that goes into it there's obviously there's a a financial standpoint there's a status standpoint establishing yourself in the industry as very serious and taking your job seriously there's learning a new architecture trying to establish a foundation for later and then there's recognition from your peers or from your counterparts in the companies in the industry that's also very important and highly valuable so it's a big risk investment to buy a component like that and it always is but you got to kind of got to put your money where your mouth is sometimes and hope it pays off so that's it for that one next question first real question is Raptor who says why does the performance of Nvidia cards scale with core count when looking within the same architecture but doesn't for AMD cards this is partially true they can 56 and 64 you don't see a lot of difference in performance once you get past 56 see use in this instance and go towards 64 AMD at this point becomes more memory bandwidth limited than anything else and Vidya tends to become Rob's limited so where you'll see and Vidya encounter issues would be something like Ghost Recon or destiny to both of these games are dx11 and both of them rely a lot on ROPS and so if you look at the titan beavers Titan XP the gains they're like 4% you look at those same two cards and something like sniper the games are way more meaningful upwards of 40% in some instances so Nvidia doesn't it's not always true that performance scale as well with core count because again Titan B's like 50 120 cores versus 3584 whatever it is on a 1080i or Titan XP yet it's 4% difference in a game like destiny or Ghost Recon depending on the settings so not always true but it is in a lot of cases same for AMD and these scales better than some games than others but AMD tends to become memory bandwidth bound before they become bound by anything else and it's also an issue of GCN reaching its limits at 64 ciues that's pretty much always been the case once you get to 4096 SP is on an AMD device you're really getting diminishing returns now first is whatever the immediate skew is below it maybe that's 60 or 56 or whatever it is so it depends a lot on architecture depends a lot of game depends on the API all that kind of stuff they become bound in different things and the for example for the longest time would become really bound in geometry and they sort of started fixing that with Polaris Polaris changed its geometry pipeline and then Vega later changed the geometry pipeline further and started adding more advanced choline tasks or sequences to it so Nvidia again it's largely ROPS I think for the most part there are some things they can become memory bound in but with g5x for example you start to become more bound by the timings of the memory than the speed itself timings are a bit looser than gddr5 but yeah alright hopefully that gives you some ideas for your question next one is Dane of Starfall I'm this one's gonna be quick in your Titan V video where you test the limit of PCIe you said that rise of the Tomb Raider is not a good benchmark title this is strange to me as I see the game on almost every GPU receive a benchmark out there can you explain why you don't think it's a good Bend for any title right as a Tomb Raider has a ton of variants you can run the test like 5 or 10 times and you'll see way more margin of error or variance than basically any other title we benchmark so we don't use it right as a Tomb Raider also it's so there's it's not really a perfect benchmarking title because you can run the built-in benchmark it has all those variance issues and other issues but if you get into actual game play it will start to smooth out a bit the problem is those numbers don't really compare to a bunch of other sites because they don't necessarily test in the same areas that's ok though so we're kind of used to that so I don't know the way I look at it is rise so when you're choosing a games benchmark you choose based on how consistent is it how easy it how easy is it for me to get to the area I want to benchmark do I want to automate it or am i testing it manually in actual gameplay so you take all those things you decide what game and then you're looking at things like API and you're looking at does this game tell me something different then games 1 through 6 if the answer is no then it's of no value to us so you look at things like DirectX 11 12 Vulcan and OpenGL ok we have a bunch of API to represent then you look at Unreal Engine CryEngine unity and whatever else maybe they're often things like frostbite or standalone one-off engines you could try to pick one of each of those and one of each API and they all represent something different so the point doesn't to have a lot of games it's having a set of very consistent games that represent different things there's no point in having 25 games if at the end of the day you could just kind of have five and it would represent the exact same thing so basically with rise of the Tomb Raider it's not very consistent so that I don't trust the results and that causes a lot of stress and retesting so waste time it's not heavily played so it's not worth the inconsistency a game like pop G I could probably argue for it even though it's pretty inconsistent and not optimized but that's because a ton of people play rise of the Tomb Raider doesn't have that going for it so not a huge reason to to fight against the title for that but yeah I mean we we basically looked to benchmark Vulcan look to benchmark compute tasks you look to benchmark things that are geometry geometrically complex things that are tessellation complex or heavy things that are using asynchronous workloads and pipelines and then different api's and you should end up with six to eight games I like to pull about eight and then we start cutting games as they prove to be bad for testing normally from inconsistency or maybe they've just released a patch that bachas all of our numbers and we'd have to rerun it so you just you know we'll cut this one out for this benchmark and we'll reintroduce it later I hope for that gives that gives kind of a bunch of different things to think about nori SS says GN covers performance engineering and stability characteristics what about warranty support what happens if something goes bad we actually wrote years ago I wrote a warranty support table that kind of gave an idea of the warranty period and whether they do cross shipping and things like that so the trouble is with warranties it's kind of different in each region I know how EVGA is warranty works in the US I know they're probably at least 90% of the time gonna be pretty good about their warranty in the US but I've heard that they're not so great for some other countries and it's just a matter of where does the company operate do they service each region the same the answer is normally no no one really has that many resources to spare so if I were to start talking about warranties they'd have to always have a big tag of like in the US and 55 percent of our viewers are not in the US so I just leave it out and honestly there are some things that although it would be it would be great if we can reasonably talk about warranties there are some things that are better left to user reviews just by the nature of aggregating thousands of them and I think warranties are one of those things because you can't necessarily trust a user review on Amazon or Newegg of a router because they're gonna say this router sucks my internet slow no your ISP is slow the router doesn't suck your ISP sucks but you can trust user reviews generally for things like warranties because that's a very basic human interaction between a company and a customer so I would rather leave it to the aggregate of thousands of users then we've done kind of like undercover warranty or service calls before with sis I'm a power cyber power origin I've called each of them basically disguised as a customer and asked for support and included those in those reviews but that's about the most we can do I can't really test a warranty process unless I were willing to like delay the review by weeks and intentionally break components see what they did or something and that's still only really one representation and then if they figure out it's me obviously they're gonna make sure everything is perfect because they know it's going to go into a review next question supa says in your review of the Fraxel define r6 while discussing GP temperatures in the vertical orientation you stated quote the vertical GB amount caused a huge temperature increase as usual can you explain a wide vertical mounted GP is tend to run hotter quote as usual we did this in a couple of videos and I didn't do it again in the r6 video the short of it is you're putting an open-face card an inch or two away from an insulating four millimeter thick glass panel those open-face cards they need air they need a lot of air in front of them to pull air into the fans and then when they're done with the air you kind of think about how they're laid out that error it can't go straight through the PCB so it has to come out the sides so it's coming out wherever the fins are oriented typically at the top and as that error spews out of the open face card it's going to go two places one it will go up into the CPU radiator if you have one above it and that'll make your sea view a bit warmer but it'll get rid of the heat and push it out of the case the other one is it'll go down hit a power supply shroud which is now probably right below your GPU because it's vertically mounted and kind of get stuck there maybe even sort of bounce back up and get recirculated and then ultimately what happens is that piece of glass starts warming up your air gets stuck against it you can't get new air in fast enough so the card runs a lot warmer sometimes in the 90s whereas a horizontal mount might be an upper 70s or the low 80s and it's just a matter of what kind of card it is now this is a scenario where so first of all full disclosure these are really meant for liquid cooling but this is a scenario where a blower card might actually be better because the blower card doesn't need as much exposure to air in terms of a big wall of air right in front of the cooler and it pushes air only in one place and that's out of the case rather than all over the case so that's what it comes down to the next one serpent XF xsf this was a fun question what was your first ever graphics card why did you get into the computer industry and reviews of computer components do you have any IT certifications bridge graphics card I think was the ATI X 1800 512 megabyte card I pretty sure there was a 256 model also I spent a lot more money on the 512 and I remember the guy at comp USA saying you're never going to need 512 megabytes of memory huh and I remember hearing the same thing when I bought a GTX 8800 Ultra later I don't remember how much memory that had but it was another one of those you're never gonna need that much and look where we are today at the time it was kind of accurate but so I out of interest I went and looked up this card I found reviews of the ATI X 1800 to see what people said about it when it came out around 2000 four or five I think this was from C night it was just the top result so this is this is just kind of fun you'll see the parallels in a second CNET said although we weren't thrilled with its design and wish that it brought more to the table than simply catching up with n videos feature set the ATI Radeon X 1800 X T's performance is hard to knock at we used for certain games the results remain consistent with the latest generations conclusion that ATI fares better at direct3d based titles like half-life 2 and Nvidia dominates an OpenGL 3d api's with doom 3 so a bit of the inverse and these better with with doom right now and NVIDIA is back to dominating with DirectX 11 so that was kind of fun to read they also said two of our biggest gripes have to do with the Radeon X 1800 XTS design the most obvious is its double-wide form factor CNET was annoyed that the card took two expansion slots I have a shelf of video cards that are 2.5 to 3 expansion slots now so that's that's the state of the industry when I got into it next or last question here Carl says can I use that GN anti-static mod matte with correct capitalization included as an oversized mousepad is it spill resistant so technically in the product page there's a line in there that a lot of you have noticed that says not intended for use as a mousepad and after that live stream where we did it overclocking for like two and a half hours I was using mice on this pad the entire time at the end of it I was like maybe undersold its potential there you can use it as a mousepad we didn't advertise it that way because I don't want people who want a mousepad buying an anti-static mat that is not really meant for that purpose if you're only planning to use it as a mousepad we appreciate the sales but you know it's just like it's there's an intended purpose for it you can use it as a mousepad but it is kind of a grippy or surface it's more towards the if you ever buy one of those aluminum mouse pads that's got a smooth and a rough surface it's kind of in between those there's a bit more texture to it so there will be some drag on the mouse so depending if you like that or not you can use it as a mousepad but I didn't want people buy it I would rather people be pleasantly surprised that they can use it for a mouse then people be disappointed that they can't use it for a mouse so that's why I didn't really advertise it that way and is it spill resistant the answer is yes we spilled some liquid on it the other day and it basically just sits there on the surface so but that was the whole point of going with this as opposed like a felt or a fabric top mousepad because this is a rubberized surface its textured and it's again it's anti-static it makes it spill resistant makes it pretty damage resistant and protects the table but not necessarily perfect for a mouse pads those those things don't go together you could use it but you know I would suggest only doing that if you also have another purpose for it like wanting to use the anti-static part of it and build stuff on the surface and that's all for this one so I think all these questions came from discord this time because the youtube comment section was so flooded with non questions that I had hard time finding on e if you have questions post them below and everyone helped float those to the top if you want to join the discord you can go to patreon.com/scishow and donate to join our discord for the community or you can always have that store that gamers Nexus dotnet or store that gamers Nexus dotnet slash mod mat if you want to buy one of these now that we've done the marketing pitch saying that it is hydrophobic or something it's not actually hydrophobic but it's kind of spill resistant I'm not very good at marketing all right subscribe for more I'll see you all next time oh that was a good one