Pages: [1] :: one page |
|
Author |
Thread Statistics | Show CCP posts - 1 post(s) |
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.15 13:36:00 -
[1]
Edited by: Zora''e on 15/02/2011 13:37:39 I have an NVidia 430GT 1 gig card, latest drivers (I just got the card a week ago, and downloaded newest drivers from NVidia). Windows XPPro (32 bit), 3 gigs ram, AMD dual core (yeah I know, I need to upgrade). My problem is that after about 3 hours of being in game, regardless of what I do (from doing nothing at all except sitting in a station, to other normal playing activities), my graphics start getting laggy/choppy.
By laggy/choppy I mean this. When I first log in, I can spin around my ship smoothly and without any discernible lag of any kind. However after a few hours, doing the same thing and it is almost a stop motion effect. This is on all graphics settings from highest to lowest.
I've been monitoring my GPU temperatures and one client (on low settings) causes the GPU temps to go from 40c to 68c. On higher settings it goes up to 74c. Multiple clients (2-3) on low settings and the temps go to the upper 80'sC. Cards rated at 94 or 98c but even so. seems to be running excessively hot to me even with just one client running.
I'm wondering if anybody else has experienced this type of graphics issue and heating with the 430GT and if so, were you able to come up with any fixes?
I did manage to bring my temps down by adding 2 more cooling fans to my box but even so, the card seems to be running rather hot for a game. -
|
|
CCP Navigator
C C P C C P Alliance
|
Posted - 2011.02.15 13:47:00 -
[2]
Moved from General Discussion to Issues & Workarounds.
Navigator Senior Community Representative CCP Hf, EVE Online
|
|
Christopher AET
Segmentum Solar The Jagged Alliance
|
Posted - 2011.02.15 13:48:00 -
[3]
yeah since some of the recent patches my card seems to be getting hot but apparently to a lesser degree. I have solved my problem with additional cooling but seems yours is somewhat more severe and I don't know what to suggest.
Also OOPE would have been the right place for this thread.
Fly safe.
Chris
|
RaTTuS
BIG Majesta Empire
|
Posted - 2011.02.15 13:48:00 -
[4]
on my 420 no problem have you present interval one set or immediate ? --
Join BIG
|
Neamus
|
Posted - 2011.02.15 13:54:00 -
[5]
The GT430 has a thermal limit of 98ŚC, so upper 80's isn't going to kill it. But yes, its a budget card and you're pushing it pretty hard with multiple clients. More case fans, or a better card perhaps? What ever you can afford really..
I'm not sure on the memory leak, EvE has a history of them tbh but I don't know whether there is a current one or not. Just stating the obvious here I know, but you'll need to start a single client and then monitor its memory usage over a period of time...
Questions: Does the lag start more quickly with multiple clients running as opposed to one? And when the lag starts, if you close the client(s) and immediately re-open them, does the lag start straight away again? Or does it then take another 3 hours before it returns?
|
Zagdul
Gallente Shadowed Command Fatal Ascension
|
Posted - 2011.02.15 14:06:00 -
[6]
Edited by: Zagdul on 15/02/2011 14:07:19 Go to your graphics settings and make sure that the interval is not set to immediate. Set it to one or default.
The "memory leak" is probably because you have resource cache enabled (this got my EVE up to 2gb once).
|
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.15 14:10:00 -
[7]
Edited by: Zora''e on 15/02/2011 14:13:52
Originally by: Neamus The GT430 has a thermal limit of 98ŚC, so upper 80's isn't going to kill it. But yes, its a budget card and you're pushing it pretty hard with multiple clients. More case fans, or a better card perhaps? What ever you can afford really..
I'm not sure on the memory leak, EvE has a history of them tbh but I don't know whether there is a current one or not. Just stating the obvious here I know, but you'll need to start a single client and then monitor its memory usage over a period of time...
Questions: Does the lag start more quickly with multiple clients running as opposed to one? And when the lag starts, if you close the client(s) and immediately re-open them, does the lag start straight away again? Or does it then take another 3 hours before it returns?
I usually run 2 clients at minimal settings. The card sits at a solid 40c until I turn eve on, then goes straight up to 68c when the first client logs in. However, I have tested my issue out with 1 client only, and then again later with 2 clients and after aprox 3 hours, the graphics start getting laggy. Doesn't matter if it's one or two clients running.
The only way to eliminate the lag is to actually reboot the computer. If I close the clients out and restart them w/o a reboot, they start up again with the choppy laggy graphics.
I actually added 2 additional case fans and managed to bring the temp down from 68c to 65c for one client. Opening the case gives me an additional 2-3 degree drop for a single client.
As for the GT 430 being a budget card... I just replaced a 512 meg 9500GT that ran two clients w/o any lag or overheating at all, ever, on minimal settings. The 430 WAS an upgrade (lol). I honestly can't see how i'm pushing a faster card with twice the ram "pretty hard", when a slower, older card ran 2 clients w/o an issue at all.
Originally by: Zagdul Edited by: Zagdul on 15/02/2011 14:07:19 Go to your graphics settings and make sure that the interval is not set to immediate. Set it to one or default.
The "memory leak" is probably because you have resource cache enabled (this got my EVE up to 2gb once).
Graphics are set to interval default (always have been) and EVERYTHING is turned to minimal settings except in the effects box, only thing off there is camera shake. -
|
Desmont McCallock
|
Posted - 2011.02.15 14:13:00 -
[8]
Separate the cooling fan from the card's GPU and apply a fresh layer of thermal paste (can be bought in any tech store). Reattach the cooling fan and you are good to go.
I've been dealing with computers around 3 decades now and most heating problems on GPU's are caused from old dried out thermal paste. Some things sit on the store shelfs to long.
|
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.15 14:25:00 -
[9]
Originally by: Desmont McCallock Separate the cooling fan from the card's GPU and apply a fresh layer of thermal paste (can be bought in any tech store). Reattach the cooling fan and you are good to go.
I've been dealing with computers around 3 decades now and most heating problems on GPU's are caused from old dried out thermal paste. Some things sit on the store shelfs to long.
Just replaced my wifes CPU fan when we thought it was starting to die (the rattle was her GPU fan though so we got me the 430 and I gave her my 9500gt to replace her 9500gt with the failing fan) so I have some new thermal paste. I'll take a look at the card and see if that is possible to do w/o borking the card. -
|
Neamus
|
Posted - 2011.02.15 14:26:00 -
[10]
So the lag issues started straight after you upgraded your card and RAM?
|
|
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.15 14:28:00 -
[11]
Edited by: Zora''e on 15/02/2011 14:35:18 Edited by: Zora''e on 15/02/2011 14:32:04
Originally by: Neamus So the lag issues started straight after you upgraded your card and RAM?
Started with the new card. I went from a 512 meg graphics card (9500gt) to the 1 gig Gt430. The choppy graphics started with the 430. Ram on my computer (the MB) is still the same 3 gigs it has been.
However, between the time I put the card in and now, there has been an EVE update (not the latest, but the one before it) and I don't remember if the lag was there prior to it, or showed up after it. Come to think about it, I think the cards been in a cpl weeks actually.
*edit* I just checked, i've had the card in since late January/early Feb, but I don't remember exactly when it went in other than that. -
|
Bal'Ayle
|
Posted - 2011.02.15 14:36:00 -
[12]
this might sound silly but i thought id throw it in there, have you checked to see if the card is using up too much power? or for that matter did you upgrade your PSU (power supply unit) when you changed cards / check it can handle it? while it may be within operational levels running two clients ups the load considerably. are you also using two monitors? some cards do not support due monitors as well as others depending on what kind of rendering the card uses
as a side, check your motherboard can it run two graphics cards at the same time? (sometimes they dont need to be the same type of card)
again check the fans on other components in your system, while fans might be running some might need cleaning or tightening since you took out and replaced the old card, power cables might not be in 100% reducing capacity of the unit etc etc. similarly check that the card you are using has upgraded capabilities some cards though older and not necessarily as "good" as other cards come built for certain things.
two gaming cards with differnt values dont pair up all the time, extra memory isnt an end all win all solution to poor graphics capabilities.
and rather stupidly what ddr is your cards memory at compared to the on-board ram? iv heard some stories that sometimes different speeds can cause hang ups as the card tried to keep up with faster computer ram etc
|
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.15 14:56:00 -
[13]
Edited by: Zora''e on 15/02/2011 15:15:45 Edited by: Zora''e on 15/02/2011 15:09:15
Originally by: Bal'Ayle this might sound silly but i thought id throw it in there, have you checked to see if the card is using up too much power? or for that matter did you upgrade your PSU (power supply unit) when you changed cards / check it can handle it? while it may be within operational levels running two clients ups the load considerably. are you also using two monitors? some cards do not support due monitors as well as others depending on what kind of rendering the card uses
as a side, check your motherboard can it run two graphics cards at the same time? (sometimes they dont need to be the same type of card)
again check the fans on other components in your system, while fans might be running some might need cleaning or tightening since you took out and replaced the old card, power cables might not be in 100% reducing capacity of the unit etc etc. similarly check that the card you are using has upgraded capabilities some cards though older and not necessarily as "good" as other cards come built for certain things.
two gaming cards with differnt values dont pair up all the time, extra memory isnt an end all win all solution to poor graphics capabilities.
and rather stupidly what ddr is your cards memory at compared to the on-board ram? iv heard some stories that sometimes different speeds can cause hang ups as the card tried to keep up with faster computer ram etc
PSU is well over the cards needs (card 300 watt, my PSU 500 Watt). Single Monitor/single card I cleaned the interior of the case, PSU and fans prior to installing the GT 430 as I had everything disconnected/unplugged and figured no better time than the present (lol) *rechecked* Connections are tight and secure/clean Cards RAM is DDR3 computers RAM is DDR2 (OCZ Platinum 1024MB PC6400 DDR2 800MHz x3)
----System Specs----
Windows XP Pro (32) Asus M2N-SLI nForce 560 Socket AM2 Motherboard AMD Athlon 64 X2 5000+ AM2 CPU Black Edition ZEROtherm Butterfly CPU Cooler BTF90 775/754/939/A OCZ Platinum 1024MB PC6400 DDR2 800MHz 3x1GB EVGA 01G-P3-1430-LR GeForce GT 430 Video Card - 1024MB GDDR3, PCI-Express 2.0, DVI, HDMI, VGA DirectX 11 Creative Labs Sound Blaster Audigy SE (cheap sound card for playing EVE, I use headphones and have no speakers hooked to the computer so it works [on board sound borked up and I decided to drop it and go with a cheap card] lol). -
|
Neamus
|
Posted - 2011.02.15 15:27:00 -
[14]
Oh sorry, you meant frame buffer memory...
When you close the eve clients down, check to see that the processes have actually stopped running.
If the problem started with your new card then you're right to be suspicious of it. It's a fermi so it shouldn't struggle with eve at all really - but being a 1st gen fermi probably also accounts for the high temps.
You might try an older driver to see if that makes any difference..? Since its a new card though personally I'd be tempted RMA it for a replacement or maybe even a different model, that way you can at least eliminate the card its self.
|
Kile Kitmoore
|
Posted - 2011.02.15 15:34:00 -
[15]
Kind of experiencing a similar issue with my ATI 5770, two clients. I have maxed out all settings on both clients however. Where I really notice problems is two clients and doing PI. My temps don't go beyond 75 degrees with two clients and XBMC running on a second display (TV) with 95% activity.
Couple of things to try: 1. Open the case and run two clients and monitor your GPU temperatures. If you see a substantial difference you have a pretty big airflow issue. 2. Is your GPU fan set to auto? Try manually adjusting. Use the NVidia CP or the following tool should help: MSI Afterburner
Good luck.
|
Bal'Ayle
|
Posted - 2011.02.15 15:42:00 -
[16]
hmm the only thing if all of that is true then the only thing looking at that is the case where the computer is ddr2 and the card is working faster on ddr3, what was your old card at may i ask?
if its not this and not the thermal paste (always worth checking as thermal damage is irreparable and will always progress to a worsening degree) then i would test another game and see if you get the same results and if so take it back to the store you bough it from and complain.
it might be worth writing to the makers of the card and asking them directly as sometimes this can net you free stuff if you play the victim card just right :D
|
Funk Engine
Gallente GoonWaffe Goonswarm Federation
|
Posted - 2011.02.15 16:11:00 -
[17]
Originally by: Zora'e Edited by: Zora''e on 15/02/2011 15:15:45 Edited by: Zora''e on 15/02/2011 15:09:15
PSU is well over the cards needs (card 300 watt, my PSU 500 Watt). Single Monitor/single card
Depends on how you read it. If you read that as the card requires 300w (Anyone remember the HD2900XT? Used ~285w itself.) then your PS is being overloaded. Meaning it only has 200w left to run all your drives and the cpu/ram/mobo from.
TBH 500w is a bit on the small side for a gaming rig anyway... Especially if its a cheapo supply with only ONE 12v rail.
|
Zora'e
Amarr Nocturnal Dementia
|
Posted - 2011.02.16 03:23:00 -
[18]
Originally by: Funk Engine
Originally by: Zora'e Edited by: Zora''e on 15/02/2011 15:15:45 Edited by: Zora''e on 15/02/2011 15:09:15
PSU is well over the cards needs (card 300 watt, my PSU 500 Watt). Single Monitor/single card
Depends on how you read it. If you read that as the card requires 300w (Anyone remember the HD2900XT? Used ~285w itself.) then your PS is being overloaded. Meaning it only has 200w left to run all your drives and the cpu/ram/mobo from.
TBH 500w is a bit on the small side for a gaming rig anyway... Especially if its a cheapo supply with only ONE 12v rail.
Yeah I checked that out myself to see if it was the issue. Card runs off the PCIe2 slot and doesn't plug into anything else. I have 3 cooling fans that are running but added up everything and doesn't seem to be the issue. However, just to be on the safe side i've ordered another PSU (Apevia ATX-WA1100W Warlock 1100-Watt ATX Power Supply - SLI and Crossfire Ready, 135mm Fan, Clear Sides) Never hurts to have additional power.
-
|
|
|
|
Pages: [1] :: one page |
First page | Previous page | Next page | Last page |