please get rid of the scaleform ui menus!

Discussion in 'Player Support' started by BlackDove, May 15, 2014.

  1. BlackDove

    No not this. If you dont understand why you shouldnt run synthetic workloads that load your GPU beyond 100% TDP then this thread is beyond your understanding and youd need to go read how these things work instead of commenting "oh you need to take care of your hardware" or some nonsense like that.

    Look up capacitors and VRMs. They dont like being overloaded even if you have some fans in your PC and think its fine.

    I have a well ventilated case and all high quality components. I even have a Seasonic X650 with an APC BackUPS XS1000 battery backup to protect it. My Gigabyte GPU with a massive cooler doesnt get above 70C because of how well i take care of it.

    Do you think all the people with issues in the 2D menus or map dont take care of their PCs? Pretty big assumption to make.

    Now with a stock fan curve and 100% power target go sit in the map or character select or store screen. Lets see how long it lasts.

    What youre actually saying is "i dont care to underatand how this works and i think its fine because i dont underatand it so ill claim that some guy who does understand it doesnt take care of his stuff". Thanks for proving my point for me as to why they need to fix the game code. Not many people actually get it.
    • Up x 1
  2. Dragam

    BlackDove : I see that by power target, you mean power limit... in that case you are right, and i agree with everything you said :)

    But i dont understand why you set it to 80% though? As long as it doesnt go beyound 100% TDP, there should be no worries. Normally i would use a modded bios that gives me unlimited power, to get the best overclocking results, but that doesnt Work too well in planetside 2, so i instead just have the power limit at 100%, to prevent it from going crazy.

    I once hit 160 % TDP in planetside 2, and gpu's reached 88c ! It should be said that i was using a modded bios with unlimited power, overclocked gpu's as far as they will go, and using super sampling and no v-sync... so it pretty much couldnt use anymore power than that.

    Keeping it at 100% power limit, the gpu's only reach 72c :)
  3. Lavans

    I guess SoE doesn't get it either :rolleyes:
  4. BlackDove

    Octiceps might be right. Its not so much a matter of them not getting it as it not being a priority.

    And did they REALLY address the performance issues with OMFG for example?

    Nope. Game only plays a little better than before. Even more unstable without PhysX and the 1990s 2D sprite particles cause framerate drops worse than PhysX did.

    Black screen driver crash happened only after OMFG as well.

    Game stiill loads peoples GPUs to 145% TDP because of the menus.

    What did they mostly do? Removed details like "extra" things in biolabs and cool things like the mech on Indar.

    Notice all the performance issue threads about the map lagging? All the overheating threads? All the "game runs terribly NO MATTER GRAPHICS SETTINGS"?

    Why would the graphics detail settings not matter? CPU beimg the limiting factor is part of it. But its also the fact that the HUD and menus are still running the same degree of synthetic load no matter how low you turn it.

    Maybe they get it and they just dont care.

    Someone who actually claims its ok to constantly load a GPU in a way that Furmark cant even do now doesnt get it.

    My GPU is nice and cool. But look at all the threads on this forum of peoples who arent.

    My game runs fine when its not broken because of a bad patch that causes it to crash to desktop. Im still usimg 314.22 drivers and i never change my setup so i can easily tell when its a bad patch for me. How many people are on here posting about bad performance with higher performance setups than me?
    • Up x 1
  5. sagolsun

    If the manufacturer put cheap components on, sucks to be 'em because they'll be replacing that card every couple of months. A properly engineered card will handle, at stock frequencies, 24/7 bitcoin mining, furmark, pick your poison. Anything less is either defective design or deliberate sabotage.

    I recommend ASUS - the caps on their mid-range and up models have double the MBTF of reference boards, excellent chokes and mosfets, voltage is rock solid, I've seen a test somewhere showing that a heavily OCd ASUS has less ripple than the reference board on stock frequencies.

    And oh, as a general rule the cooling is way overengineered and would probably support a card with double the TDP.

    The "VRM killer" is a BS excuse for releasing defective products and placing sub-par components at the consumer's expense. An excuse manufacturers just LOVE to hear coming from people's mouths.
    • Up x 1
  6. BlackDove

    Please explain how constantly unnecessarily loading a processor and its power components beyond its TDP because of a power virus type code is acceptable?

    I get what youre saying about buying good quality power components. I do buy them. You completely missed the point.

    What youre saying is basically "oh its fine to drive around in first gear redlining the engine. If your expensive high quality car cant take it then it sucks."

    Doesnt matter how nice or well made it is. You still shouldnt excessively load it for no reason.

    Lazy people love to hear whats coming out of your mouth. "Bad code? Power virus? If your PC cant take it it must be junk!"

    I recommend Asus and Gigabyte for the same reason you do, but i dont intentionally want my card sitting at 145% TDP because of a badly coded 2D menu.

    Not to mention the performance bebefit for everyone if the most demanding thing in the game starts being the 3D graphics instead of the 2D menus!
  7. Lavans

    There's something you don't realize - 100% GPU usage & higher than 100% TDP =/= power virus. Modern GPUs have dynamic TDP based on heat/usage. Any card with driver defaults will have fluxuating TDP, because that's what AMD and Nvidia designed them to do. A GPU is fully loaded at 100% with a TDP of 100% (or more) does not confirm the existence of a power virus.

    Stop pretending you know something that the SoE devs don't. If PS2, or any component of PS2, acted like a power virus, fixing it would be a priority. But it's not a priority, because there is no "power virus". Do you notice the forums booming with people complaining about their GPUs overheating? Because I sure don't.
    • Up x 2
  8. BlackDove

    Im pretty sure i am the one who just explained how GPU boost works... Thanks for restating it.

    Anyway i never said that greater than 100% TDP is ALWAYS a power virus. A power virus is an atypical workload that loads the GPU to beyond its TDP without actually doing anything that SHOULD load it in such a way.

    Should a simplistic map or 2D flash based HUD load modern GPUs yo 100% load and GREATER than 100% TDP?

    WHY would a 2D menu or HUD cause such a load? Answer that and stop making excuses.

    The forum has been filled with people having issues with the map and menus since launch. Here are just a few. "My gtx 580 likes to run hot whenever there is nothing on the screen. Map, loading, warpgate, looking a wall and such and such. It doesnt run nearly as hot during fights however."

    Do you notice those?
    • Up x 1
  9. Lavans

    A small handful of people out of thousands over the course 1.5 - 2 years, complaining that their components are running hotter than -they- expect, does not imply a software related issue.

    You perceive there to be an issue because you want there to be an issue. But there's not. SoE is not going to let their own product be a notorious hazard towards other people's computers.
    • Up x 2
  10. NaySayer

    I love it when 1 out of the thousands of people that don't have or care for this problem say it isn't the game, it's the thousands.
    • Up x 1
  11. BlackDove

    Lol. You really think its just a few people when most of the general tech support section is filled with people either crashing to desktop with no error message or performance thats lower than expected on machines that can play games with better graphics more easily.

    I want there to be an issue? Ok what motive do i have to want the menus to excessively load peoples GPUs? Now youre resorting to ad hominem because you cant successfully prove your point.

    Even people who have well cooled PCs like i do dont want to load them to 145% TDP FOR A 2D MENU. The menus and HUD shouldnt be the most demanding thing in a modern game, implying that theyre badly coded if they are!

    What is your motive for claiming that a games menu which presents a load that not even OCCT or Furmark can produce is ok?
    • Up x 1
  12. Lavans

    What proof do you have that PS2 is a power virus? :rolleyes:
    • Up x 1
  13. BlackDove

    The 2D menus, map, store and HUD load the GPU like a power virus. The game itself is not a power virus.

    A 2D menu or other simplistic element like that should not present a load that gets a GPU to 145% TDP. A 2D menu is not a difficult workload for a GPU unless it includes code in it which saturates a GPU with unnecessary calculations because its improperly coded.

    Its like youre making an argument to KEEP a bug in the game. Why keep 2D elements in a game that load it more than the ACTUAL 3D GAME elements do?

    Are you confused between PS2 and its 2D menus? I dont think you are but you are trying to misquote me to avoid answering the question and successfully defending your point.

    Perhaps its because you know you cant.

    All those threads and the rest of the threads where people describe how badly it performs are evidence.

    Wheres your evidence to back up your claim? Didnt think so.
    • Up x 2
  14. Lavans

    No they don't. Try again.

    • Up x 1
  15. BlackDove

    Actually your picture illustrates the behavior described in the threads where people have issues perfectly!

    Why is the map alone at 35% power in P state 2 at 810MHz .987V 15% MCU load while the store is ALMOST DOUBLE at 68% power in P state 0 now, boosting above your default BOOST clock to 1097MHz 1.175V 28% MCU load?

    LOL you even disabled automatic fan control between pictures. Why? Did you think people would fall for that?

    Gonna tell us that the store page which is about as complex as a flash based website requires the GPU to boost beyond its default boost clock, double its power usage, double its memory controller usage and increase its Voltage that much AND change P states(and apparently stop the fan speed from ramping lol)?
    • Up x 1
  16. Lavans

    I love how you first define a power virus as a piece of software that causes a component to exceed its TDP. Yet, now you retract your statement and rephrase it to make it convenient for your sake of the argument.

    You do realize the game continues to render behind the store, but not the map, right? That would be a clear indication as to why the store is drawing more power where the map is drawing less power. Derp.

    Also, do you even know what MCU is? Because the way you reference it, you make it sound like you have no idea what it measures.

    Maybe because I have EVGA loaded in the background with custom fan controls, and launching Inspector conflicts with the setting upon launch? Notice how I'm in the middle of a zerg in the first picture, but not in the second? Maybe I closed Inspector to get it out of the way to participate in the zerg and launched it later when it was safe. Regardless, why does it matter?
  17. BlackDove

    So youre not really controlling your tests very well and you have multiple overclocking utilities open with one hidden and one selectively shown?

    How convenient.

    And in Nvidia Inspector it refers to the memory controller.

    And in any case its pretty bad design to have everything rendering behind the store.
  18. Lavans

    I don't "control" my tests. I play the game. Simple
    EVGA doesn't show as much detailed information as Inspector, and as such, serves no purpose in being in the foreground. Also simple.

    Wrong. It refers to Memory Controller Usage. It measures the amount of data which loads into VRAM being passed through the framebuffer.

    Then suggest to SoE to give the store a solid backdrop so it culls the 3d rendering and reduces load.
    • Up x 1
  19. LibertyRevolution

    Seems a lot of people in this thread seem to be thinking you are complaining about in game temp..
    We are not talking about in game temps, we are talking about when you look at map/terminal/profile screens, your temps spike.

    So in game 65ºC at 66% fan speed.. In Menu 76º at 85% fan speed...

    What happens if you put it on a terminal and walk away for 5mins??
    It is hitting 90ºC with 80%+ GPU use. This is with 85% fan speed... as in like loud as a hair dryer.

    Being on any of these menus will work your GPU harder than even the largest in game battle...

    Hopefully this will clear up the confusion some have about what you are talking about.

    We did that.. in May of 2013 when I shot these videos.. a year ago..
    BlackDove and I both posted a thread on the bug forum.. a year later they are still ignoring the problem..
    Their view is it is not a problem as you shouldn't be on a menu for that long...
    • Up x 2
  20. Smagjus

    Please keep in mind that Lavans tries to prove that the UI is not a power virus like suggested in the beginning. That it is inefficient doesn't have to be proven.

    The screenshots are solely meant as indication that this may be not the case.

    I also wonder, although it was a little longer ago, if according to you, Blackdove, a GPU is overheating at 70°C or if this significantly reduces its lifespan. Because the arguments that you made in the beginning let it seem like this was your stance.

Share This Page