Comments by "" (@Wingnut353) on "Gamers Nexus" channel.

  1. I'm not sure why anyone thinks AMD's shenanigans with HTATI has anything to do with x86 licensing which nobody cares about in china... only if it shipped to the USA or other country that cares about our licensing. It's about AMD has the IP and does the development and they can just sit back and hit print on the CPUs essentially redeveloping x86 to the level of Intel or AMD independently is cost prohibitive... The joint venture part is to get around the USA blocking export of tech... so a chinese company has to actually do the manufacturing (and honestly it still seems a bit iffy), also the entire reason those CPUs won't show up through normal channels here in the USA is because of Intel licensing agreements.
    1
  2. AMD confirmed that Renoir on desktop has 24PCIe lanes (12 on picasso).... as soon as that is available it will pretty much stomp any intel CPU into submission due to having a new fast IMC and low latency controller, in addition to the availability of Vega + encoders etc on die. Also Nvidia buying ARM would be one of the worst things to happen to the industry in a long time.
    1
  3.  @sweetpotatox1007  THIRTY is bigger than 20, it uses more text....
    1
  4. @Gamers Nexus Don't forget that Windows 7 actually does have support for DX12 but only for select games that partnered with Microsoft to port them so it may be possible that CDPR did this. In any case they should have just ported the engine to Vulkan...
    1
  5. As a computer engineer... your dissing of the guy saying it probalby would affect longevity is out of line. The fact is you have a bell curve of console failures, that memory module running hot is almost certainly going to push that bell curve in the wrong direction. And that gets down the the bean counters again... and isntead of spending and extra buck to cool thtat chip they wimped out.
    1
  6. The problem with Navi is they changed a lot of stuff all at once with probably not ehough testing... the power managment for the chip got totally revamped (Vega 20 has it also but it also still has the old power managment). As far as the 5600XT... that's pretty obviously a discrete GPU to match with thier upcoming mobile CPUs.
    1
  7. As a computer engineer what you are saying is complete nonsense... UEFI is modular, chipset support and CPU support are basically independently developed in the BIOS. Only supporting a product for 3-4 years is considered insane in many industries.
    1
  8. Yep you did well... I'm sticking with my 1700x and x370 untill AM5 at this rate though.
    1
  9. Actually unless you OC... most x370 boards are perfectly fine. My Astrock x370 runs a 1700x perfectly fine and unless you OC or go for the 12+ cores then it doesn't matter. The real reason you might want to upgrade is if the BIOS is not stable on your board in which case you'll have to just google what others are experiencing on your board.
    1
  10.  @BeamMonsterZeus  The lan cards above 1Gbit are pointless anyway you should just get a cheaper board, an SFP+ card for $10 on ebay and some DAC cables... and let the good times roll. 2.5G-BaseT and 5G-BaseT are far too expensive to be practical you'll spend more on a basic router for it than a high end AM4 mobo...
    1
  11. Steve you should reevaluate your position on cloesd GPU code.... currently Intel and AMD both share code for thier drivers without issues as the so called IP just inst' that big of a deal since all of them are cross licensed anyway even Nvidia. Closed GPU code is so bad that HPC developers go to the trouble to develop thier own drivers to work around the BS that Nvidia pushes on them without fixes. It's very similar to the situation of AMDs drivers on windows, except AMD does have open source drivers they just aren't ported to windows, perhaps someday they will be (some of the code is shared between them though). Nvidia has no excuses for being closed, as if anyone wanted to take AMD/Intels drivers and build thier own it would be a billion dollar effort... and that is just a non starter, even Samsung is licensing AMD's GPUs outright rather than continuing to develop thier own... stealing just isn't practical. Also AMD and Intel SHARE OPTIMIZATIONS on Linux... there is no reason not to. This pushes the real optimizations down into the hardware, as all the software should function and perform the same anyway. You say bullshit like this, and then turn face when we get bullshit drivers.... you realize that its he very reason Richard Stallman (the old creep) even bothered to write open source code to begin with? He had this crap driver for for a printer at the time with closed drivers and it pissed him off... at some point he quit being productive though and became a software hermit... .but whatever that is beside the point. There are over a dozen open source driver developers at AMD at this point I believe.... making good money off open drivers... Also AMD goes as far as to run both their closed and open driver code on top of the same kernel driver, so code gets released first into the closed driver and then makes its way out into the open driver after legal review. Nvidia could do effectively the same thing, but they choose not to because they are lame.
    1
  12. He's a bit stuck in the past on that... and doesn't even run an NVMe drive in his systems. while most people buying an x570 board should be pairing it with something like a Rocket 4 1TB SSD or perhaps one from Corsair or Gigabyte which are two other companies with PCIe 4 ssds....
    1
  13.  @BeamMonsterZeus  This also is true, everything he did say about them was valid.
    1
  14. At least with intel you know you wont' be able to upgrade -_- ... and knowing a little about thow these things work... the reasoning behind dropping x370 is exactly the same reason as x470 customer hostile cash grabbing.... and I won't be party to it.
    1
  15.  @Capybarrrraaaa  Correct as that is only half as bad... alot of voodoo goes on in the network code of FPSs to make it that way, but you can't do anyting like that for input latency.
    1
  16. Watched more of the video and I have to say I agree... with one exception if you want to be fair, nvidia's vendor lock in habits (which are very MS like) need to be bashed. If you want competition there is no room for vendor lock in. If AMD doesn't have a halo card to take on the 3090... they need to fire Frank Azor... and maybe a few others....
    1
  17. Only if you have BPF JIT enabled... which is not default. Otherwise AMD isn't vulnerable to any of these.
    1
  18. Fryzen... thats one letter difference in the same market essentially as Ryzen. So would get struck down as copyright infringement.
    1
  19. There is no one else to compete with Intel/AMD is not right.. IBM Power 9 is right up there and can trade blows with AMD/Intel in supercomputers. There are some ARM processors in the supercomputer market that probably also trade blows due to having more local memory (HBM on package).
    1
  20. All It guarantees... is that I will in fact never be buying any upgrade on AM4 and will just wait for AM5....
    1
  21. And why is that... the chipset is nothing but a pherpherial IO controller... ram routing and PCIe wiring are not drastically different on 300 and 400 series boards and the VRMs on 300 series boards are pretty much fine for all the low to mid range CPUs expect perhaps some boards might not handle the 12-16 cores.
    1
  22.  @fortunefed8719  The Hitachi DeathStar
    1
  23. To be fair.. the 7nm parts aren't even that power hungry... an 8 core Zen 3 probably uses less power than an 8 core Zen 1...
    1
  24. Yes... they have done this since forever and sometimes there have been compatibility issues with some cards. I've seen some servers that offer PCIe 1 2 and 3 rates etc... also.
    1
  25. Something is wrong with the video in this stream? I can play other 1080P60 video streams just fine but yours is just stuttering and hanging all over the place??? Vega FE GPU... maybe something there dunno. Edit also... looks like this video defaulted to AVC1... I think perhaps it was choking on the 60FPS AVC1.... blocking 60fps made it work OK.
    1
  26.  Interlace  Updated from 19.12.2 to 20.1.2 AMD graphics drivers and it seems OK now? I don't have many addons besides videodownloadhelper and ublock. Also I use Firefox exclusively.
    1
  27. x570 Taichi has 3 M.2 slots... slot three disables the bottom 4x PCIe slot though and it isn't wired to sata but big woop. I just recommended it to a friend as a probably the best all around good board on x570/AM4 and it has 8 SATA ports which is good for him since he wants to expand storage in the future.
    1
  28. Read part manuals and google part numbers for a few years and you will be just like buildzoid :P
    1
  29. From what I can tell and even Buildziod has said as well... there is not a ton of point to any of the boards above the Asrock x570 Taichi's $250 price point *unless you have a specific need*.
    1
  30. Undervolting?
    1
  31. I was thinking like dang BZ is bumpin his WPM up.... then I realized he was running at 1.25x
    1
  32. Like... antivirus software, they largely do file IO and querying files and kernel information, so you will probably see you're already slow antivirus slow down even more. I wouldn't be supprised if the very popular WinDirStat seens a huge performance hit...
    1
  33.  @010203109  Eh add a USB/SATA combo card... and it will acutal put one of your PCIe 1x ports to good use. https://www.ebay.com/itm/Power-eSATA-II-USB-3-0-HUB-Combo-PCI-E-Card-with-Motherboard-20-pin-Connector/192881116444?hash=item2ce89c451c:g:4RAAAOSw3v5Yo~28
    1
  34.  @ActuallyHardcoreOverclocking  Well dang, I guess it must have been like you said and they found that it sucked... I updated my other comments also to reflect this.
    1
  35.  @ActuallyHardcoreOverclocking  Just for reference and to any snuff out any further doubters This is a link to your photo of your retail x570 Taichi https://i.imgur.com/v6LLuTf.png
    1
  36. Hardly... its just a matter of not deleting initialization code for the x470 chipset from the BIOS... not additional work. Compared to the amount of money they would be making on selling upgrades for older boards... yeah its a trivial amount of work.
    1
  37.  @itsallgoodman8867  Well x370 isn't all bad either... at stock settings almost identical, the main difference I think is PBO may be a bit less capable on x370 but from what I remember it still works. You just have to be careful as few x370 boards have good BIOS support for Zen 2 and even fewer have BIOS flashback. If you are compromising the board, it stands to reason you will probalby have a Ryzen 5 CPU also which won't put too much strain on the VRM.
    1
  38. Steve...AMD just needs to make cards that dont suck, and the marketing will follow, it's a waste of time to market inferior products. Some things I dont care about like DLSS as its wasted silicon IMO better used for rendering higher detail natively.
    1
  39. You mean like an E-ATX board with SAS and such?
    1
  40. Explicit mGPU != SLI/CrossFire ... explicit mGPU via DX12 and Vulkan is the future. Also you'd have to be an idiot invest in coding your game in DX12.... for one platform. When you can just use Vulkan and your game and engine will run everywhere. Engines are where we'll see explicity mGPU showing up first.
    1
  41.  @GamersNexus  While this is true if you are doing a new build you would be remiss not to grab a Sabrent Rocket 4 1TB on sale for under $170... that's just a steal and would make the PCIe 4 sort of useful.
    1
  42.  @naveenbattula  DUUUDE... don't comment if you aren't going to bother understanding what I wrote. By all means bump up the CPU before going for the faster drive. Also that's a Crucial drive not corsair you are mentioning. And that drive is literally over 3x slower than the Rocket 4.... sure it's a good call if you are cash strapped or want to upgrade a different part but it's also a good point to upgrade if you have upgraded the other stuff already.
    1
  43.  @naveenbattula  That's not a good argument, it's like saying an FX series CPU is fine for gaming these days... sure it will get it done but it won't be as good and you will be missing out on significant amounts of performance even if it isn't something you actually can notice without reference. The main thing you lose on SATA SSDs is IOPS... which are important for interactivity during heavy system load... with an NVMe drive you may never notice the heavy IO occuring in some background task etc... I know people that up until last year though all SSDs were a waste of money then they acutally got an SSD in their new computer and learned better.
    1
  44.  @4fiHysteria  3x difference for 170% the price but yeah diminishing returns are a real thing. Faster IO will reduce stutter though so it can still be relevant in gaming situations. Or for people that do development etc....
    1
  45.  @naveenbattula  I mean a lot of people would disagree... all the people that have been buying faster samsung SSDs for the past several years instead of slower but cheaper ones. FPS isn't everything either, a faster SSD will be able to load shaders and textures faster than a slower drive and not everything can always be cached, when a shader is compiled it usually hits the disk and you get stutter. I do agree it isn't the same thing, but it's kind of a stupid argument, moving from a P4 to Zen 1 isn't the same as from Zen 1 to Zen 2 either.... but people are still doing exactly that incremental upgrades also if you are buying new and you don't have severe budget concerns may as well get one of the fastest SSDs on the market for not much more. It certainly makes a lot more sense than other things enthusiasts spend time and money on.
    1
  46.  @naveenbattula  well on point is that many of the cheaper NVMe drives actually have extremely poor random write as well like as bad as SATA drives. Anyway people are gonna build what they can afford or what they think is right anyways lol and like you say it isn't imperative to even have anything really faster than a SATA SSD... even if there is potentially a bit better experience to be had on faster hardware.
    1
  47.  @Jake12220  That absolutely is not true for dual cards as I said...before you went off on a tangent. Also mainstream boards dont have full 2x PCIe 16x they have 16x +8x or 8x + 8x etc... so mid you have a use for Multi gpu be it compute or graphics PCIe 4 will help inter card communication. Also these ssd is benchmarks you speak of...I bet they were ran on windows which has crap IO performance to begin with, so the reason it doesnt scale is windows has not switched to polling or hybrid IO like Linux has for low latency block devices.
    1
  48. that is an extremely gaudy card I'm I'd go as far as to say downright ugly...
    1
  49. You should cut the rad open....if it has other metals than copper in it....you get corrosion. Sure you can attempt to mix metals but it is generally a very bad idea.
    1
  50. That's 3 generations boss... Zen+ isn't a new generation its just a refresh of Zen 1.
    1