Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

Imakuni

Member
  • Content Count

    7,828
  • Joined

  • Last visited

Reputation Activity

  1. Informative
    Imakuni got a reaction from KnightSirius in Clock speed or more cores? F@h   
    Rather than 3x SLI 980, get 2x SLI 1070. While upfront cost of a 980 is smaller, power consumption will play a factor in the long term, making 1070 a much better option. Maxwell is efficient, true, but Pascal is much more.
     
    As for the CPU, an i3 will do. Due to the way the software works, you don't really need a "core" to feed an Nvidia GPU, you just need a "thread", even if HT; the CPU isn't actually doing much, it's just the driver hogging useless calls that have to be fed with a thread, not with actual usage. If you want to keep the system modern, you can go skylake; otherwise, an older platform will do just as fine.
     
    Just be aware of PCIe lane constraints. While you don't exactly need SLI on it's own, I still recommend getting an SLI capable mobo, as those have at least 8x features on 2 slots. Going lower causes a drop in performance. In fact, that's another reason why I recommend 2x 1070 rather than 3x 980: lack of bandwidth / need for a much more expensive motherboard.
  2. Informative
    Imakuni got a reaction from TopHatProductions115 in Teslas for folding   
    I bet 10$ this was entirely Sony's fault. If I was at FAH, I'd do everything possible to keep the project alive.
     
    My particular bet is that this caused increased overhead on the PSN, and thus it was shut down, but I may be WAY off on this one...
  3. Like
    Imakuni got a reaction from TopHatProductions115 in Teslas for folding   
    PS4 / Xbox One can't...
     
    I know this sounds like a joke, but it isn't. There was a time when PS3 actually folded, but that's long gone now. And it never came back for current gen.
  4. Agree
    Imakuni got a reaction from Delano.888 in 3000mhz vs 3200mhz ram, is there any point?   
    Faster RAM totally doesn't help, right?
     
    I can give you many more examples, if you want.
  5. Informative
    Imakuni got a reaction from AvogadrosDog in 3000mhz vs 3200mhz ram, is there any point?   
    EuroGamer. I can pick from other sources if you'd like, though.
    I've done it myself as well and there can be anywhere from margin of error difference to 10%.
    Right... so you pick a potato GPU such as the 660ti, put it on Ultra 4X MSAA to create the biggest bottleneck you possibly can, test on 2 games and then come the ballz to say it doesn't matter? C'mon... you can't be serious.
     
    Also, I suggest you ignore all videos from that dude, for he clearly doesn't know what he's doing.
    CPU? GPU? MSAA levels? Only 2 games? Same story as the other video.
    Er... excuse me? Do you even understand what's going on?
     
    You can see an i3 6100 moving from the standard 2133 RAM to 2666mhz; you can see the boost there. They also throw in an i3 4130 moving from standard 1600 to 2133 RAM; you can't compare it to the 6100, that much is true, but you are comparing teh 2 4130 tests to see the difference in RAM speed.
     
    Just because you don't understand a chart doesn't mean that it is false. But okay, you want youtube benchmarks? Fine, have a video from them if you don't believe the chart.
     
    Imprecise. If you have an unbalanced setup with settings cranked WAY above those that your GPU can handle, better RAM won't help. And in some games, even a balanced setup with good settings is not going to benefit from RAM. In those instances, you'd be right to say that RAM doesn't help.
     
    But generically saying that at max you get 1fps in every case is just straight up wrong. RAM does matter, and it's time for people to start realizing that.
  6. Agree
    Imakuni got a reaction from WereCat in 3000mhz vs 3200mhz ram, is there any point?   
    EuroGamer. I can pick from other sources if you'd like, though.
    I've done it myself as well and there can be anywhere from margin of error difference to 10%.
    Right... so you pick a potato GPU such as the 660ti, put it on Ultra 4X MSAA to create the biggest bottleneck you possibly can, test on 2 games and then come the ballz to say it doesn't matter? C'mon... you can't be serious.
     
    Also, I suggest you ignore all videos from that dude, for he clearly doesn't know what he's doing.
    CPU? GPU? MSAA levels? Only 2 games? Same story as the other video.
    Er... excuse me? Do you even understand what's going on?
     
    You can see an i3 6100 moving from the standard 2133 RAM to 2666mhz; you can see the boost there. They also throw in an i3 4130 moving from standard 1600 to 2133 RAM; you can't compare it to the 6100, that much is true, but you are comparing teh 2 4130 tests to see the difference in RAM speed.
     
    Just because you don't understand a chart doesn't mean that it is false. But okay, you want youtube benchmarks? Fine, have a video from them if you don't believe the chart.
     
    Imprecise. If you have an unbalanced setup with settings cranked WAY above those that your GPU can handle, better RAM won't help. And in some games, even a balanced setup with good settings is not going to benefit from RAM. In those instances, you'd be right to say that RAM doesn't help.
     
    But generically saying that at max you get 1fps in every case is just straight up wrong. RAM does matter, and it's time for people to start realizing that.
  7. Like
    Imakuni got a reaction from Lane Matzeder in How badly will streaming damage my CPU?   
    Are you a twitch partner? Because if you aren't, 1706:720 at 60fps is an extremely dumb resolution. You HAVE to lower it down, otherwise your viewers will suffer from it.
    Don't worry, that message is just saying that your CPU doesn't have enough power to encode + render whatever is it that you are gaming smoothly. It's a warning telling you that you should use lower settings to save on CPU power so that OBS and your games don't fight for limited resources, not a warning telling "omg, CPU EXPLODEZZZZ".
  8. Like
    Imakuni got a reaction from DarkSmith2 in 3000mhz vs 3200mhz ram, is there any point?   
    EuroGamer. I can pick from other sources if you'd like, though.
    I've done it myself as well and there can be anywhere from margin of error difference to 10%.
    Right... so you pick a potato GPU such as the 660ti, put it on Ultra 4X MSAA to create the biggest bottleneck you possibly can, test on 2 games and then come the ballz to say it doesn't matter? C'mon... you can't be serious.
     
    Also, I suggest you ignore all videos from that dude, for he clearly doesn't know what he's doing.
    CPU? GPU? MSAA levels? Only 2 games? Same story as the other video.
    Er... excuse me? Do you even understand what's going on?
     
    You can see an i3 6100 moving from the standard 2133 RAM to 2666mhz; you can see the boost there. They also throw in an i3 4130 moving from standard 1600 to 2133 RAM; you can't compare it to the 6100, that much is true, but you are comparing teh 2 4130 tests to see the difference in RAM speed.
     
    Just because you don't understand a chart doesn't mean that it is false. But okay, you want youtube benchmarks? Fine, have a video from them if you don't believe the chart.
     
    Imprecise. If you have an unbalanced setup with settings cranked WAY above those that your GPU can handle, better RAM won't help. And in some games, even a balanced setup with good settings is not going to benefit from RAM. In those instances, you'd be right to say that RAM doesn't help.
     
    But generically saying that at max you get 1fps in every case is just straight up wrong. RAM does matter, and it's time for people to start realizing that.
  9. Like
    Imakuni got a reaction from DarkSmith2 in 3000mhz vs 3200mhz ram, is there any point?   
    Faster RAM totally doesn't help, right?
     
    I can give you many more examples, if you want.
  10. Like
    Imakuni got a reaction from ArduinoBen in can furmark damage a gpu?   
    Bro, chill. You are fine. Forget about it, your GPU is fine. It's like parents hitting their children: a few taps won't kill them, but doing it 24/7 surely will.
     
    (Btw, fix the "eunning" typo on OP.)
  11. Agree
    Imakuni got a reaction from GDRRiley in LTX 2017 Details   
    Dear @CPotter , please fix the post for your loyal night theme users.
     
    And while you're at it, make sure to talk to Linus to put a rule in place so that all members on LTT are forced to use night theme, as to prevent this from happening ever again.
  12. Agree
    Imakuni got a reaction from Tiz in LTX 2017 Details   
    Dear @CPotter , please fix the post for your loyal night theme users.
     
    And while you're at it, make sure to talk to Linus to put a rule in place so that all members on LTT are forced to use night theme, as to prevent this from happening ever again.
  13. Funny
    Imakuni got a reaction from aL_eX in can furmark damage a gpu?   
    70ºC or less -> You are golden 70-80ºC -> You are good 80-90ºC -> You should either stop or ramp up fan speeds 90º or more -> "HELP ME, MY GPU IS ON FIRE, AND I CAN'T USE WATER ON IT!"
  14. Agree
    Imakuni got a reaction from aL_eX in can furmark damage a gpu?   
    Bro, chill. You are fine. Forget about it, your GPU is fine. It's like parents hitting their children: a few taps won't kill them, but doing it 24/7 surely will.
     
    (Btw, fix the "eunning" typo on OP.)
  15. Agree
    Imakuni got a reaction from LooneyJuice in 4790k OC is at 100 degrees!!   
    It is not an error, it's just that AIDA is a potato test while P95 is designed to squeeze as much performance as possible out of the CPU, taking advantage of features other programs don't use.
  16. Funny
    Imakuni got a reaction from done12many2 in 4790k OC is at 100 degrees!!   
    Reminder that AVX-512 is (supposedly) coming with Cannon/Coffee.
     
    Guess I'll sell my oven and use one of those processors instead...
  17. Agree
    Imakuni got a reaction from done12many2 in 4790k OC is at 100 degrees!!   
    Let's see... weak cooler, high voltage, intensive test.. yeah, 100C seems just about right.
     
    Assuming it is in fact running at 1.3v (which might not be true, seeing as "this is what I told it to run at" is just plain wrong), lower your voltage down. Try 1.2v (or lower, actually), should be much more manageable.
  18. Agree
    Imakuni got a reaction from hiran91 in Two duel channel ram kit + one single stick   
    You can still use dual channel just fine, you just have to put both 4 gb sticks in one channel and the 8 one in the other.
    Wrong. I'm on tablet right now and cant really get things easoly, but look for intel's guideLines on dual channel operation and you'll see that, in fact, 3 sticks can do dual channel, if the conditions are right.
     
     

     

  19. Agree
    Imakuni got a reaction from SpaceGhostC2C in Two duel channel ram kit + one single stick   
    You can still use dual channel just fine, you just have to put both 4 gb sticks in one channel and the 8 one in the other.
    Wrong. I'm on tablet right now and cant really get things easoly, but look for intel's guideLines on dual channel operation and you'll see that, in fact, 3 sticks can do dual channel, if the conditions are right.
     
     

     

  20. Informative
    Imakuni got a reaction from drax_mafra in ASUS Z-170a Overclocking   
    Normal behaviour, that's how electronics work.
     
    http://dlcdnet.asus.com/pub/ASUS/mb/LGA1151/Z170-A/E10611_Z170-A_UM_V2_WEB.pdf?_ga=2.92118170.1870799286.1495624592-1420393963.1487242403
     
    Page 2-25
  21. Informative
    Imakuni got a reaction from drax_mafra in ASUS Z-170a Overclocking   
    High load causes voltage to droops, there's just no way around it (can't mess with physics). If you were to put a single voltage and just provide that, it would droop. So to try and compensate for that drop and achieve the actual specified manual voltage, LLC quicks in and add some extra on top of it, so that it droops from extra to actual input when load is high.
     
    Supposedly, Auto LLC would just pick the perfect level, but that doesn't really happen in practice. And that's where the user comes in: through trial, error, and human observation, you can pick the appropriate level for your setup so that load voltage matches whatever you want.
  22. Like
    Imakuni got a reaction from JoostinOnline in DIY Ram Heatspreaders possible?   
    Oh yeah?
     
    Still, if you want to do it just for the effort of trying, you could try a couple things:
    Print a pattern / color / whatever on photo paper. Glue a bunch of sheets together (to add thinckness) and cut to size aftewards. The inside of a box of juice is kinda metalic-ish, should give a nice effect if you were to add some cellophane to it, or even just raw. Mirror paper, along with slices of the previous box juice solution (again, for thickness) could be an option as well.
  23. Like
    Imakuni got a reaction from DrMikeNZ in cant overclock g.skill memory?   
    For context, Ryzen is only supposed to do 1866 with 4 dual rank sticks, so I guess your 2133 is kind of something already...
  24. Like
    Imakuni got a reaction from Ruffleduffle in Mining, Boinc, and Folding   
    Dear @RAM555789,
     
    Please scroll down to the bottom of the page and change the theme to Night. Then, scroll back up and, once you see what happened to OP, fix it. Feel free to continue using the superior night theme or move back to day afterwards.
     
    With love, Imakuni, Master Pokemon Singer.
     
     
     
  25. Agree
    Imakuni got a reaction from alextheaverage in Prime95 anyone?   
    http://www.primegrid.com/forum_index.php
     
    Not exactly Prim95, but we might as well say so... the program used (LLR) is the exact same as P95's, only with different input values.
     
    You've got it the other way around: P95 is a prime searching program that just so happens to also have a stress test mode to assure your hardware is up to the taks of prime hunting.
     
    Oh, and btw, it's not pointless:
     
×