Jump to content
Sign in to follow this  
King Haggis

GeForce GTX280

Recommended Posts

I think SLI is the future, ...

I mean 1+1>2

I believe that this is the target and it will be achieved in the near future just with the proper drivers.

Same here.

For the current situation and situation in the past, I agree.

But for the future I believe SLI, or at least multi-GPU technology will be the way to go to. Kinda like the Multi-CPU progression.

But that doesn't change the current situation smile_o.gif

Share this post


Link to post
Share on other sites

Well, installed my new PSU today. Still having the same problems. I get a black screen after playing for about 2 minutes. Same problems with Hellgate London.

I looked it up and found the following website:

http://www.nvnews.net/vbulletin/showthread.php?t=114925

It seems I'm not the only GTX280 owner with black screen problems. I put the GX2 back in and gave me no problems, just that it gives random stutters so I don't want to keep that card.

I guess I have to RMA the card and hope the replacement functions better. Sigh... this is very frustrating. I wish I never sold my 8800 Ultra. It's a darned shame the old cards are better than the new ones.

Share this post


Link to post
Share on other sites

Try using a tool as Everest, to monitor the heat of the card during load etc.

Also, they seem to point at Powersupply quite a lot, even in that topic. I do not experience the same problems, with the PC Power & Cooling 750watt (1 12v rail of 60A)

Share this post


Link to post
Share on other sites

I have a 1000W Antec PSU now which has been given a lot of praise in reviews. It's not a single 12V rail but 4 of them giving a combined 70A (or so, I forgot). Maybe a single 12V rail is better but come on, this PSU MUST be good enough?!

I don't want to monitor my temperatures now, I have just put my GX2 back in. I'm getting a bit sick and tired of this PC. Don't take it personally, I just have to cool off this weekend. I'll take another look into it tuesday. Monday I'll order a new GTX280 and send my "old" one back. I hope I'll have a new one on tuesday. If it keeps giving the same problems then I'll have to go looking for another PSU with a single rail. SIGH!!!

confused_o.gif

I'll let you guys know, thanks for the advice.

Share this post


Link to post
Share on other sites

Well, I somehow found the energy and patience to do one last search on this subject before I give up and go away for the weekend.

I found this thread:

http://forums.guru3d.com/showthread.php?p=2735492

There are people claiming a multiple rail PSU should be working as well. If the replacement GTX280 doesn't work, I still want to give a single rail PSU a try, regardless of what those people say but right now I still doubt the GTX280 is 100% allright. If a new GTX280 and a new single 12v rail PSU don't work, I'll have to ask for a refund and get a new or second hand 8800 Ultra once again or get a new 9800GTX.

Share this post


Link to post
Share on other sites

Ah, sorry, I thought you had a 600watt OCZ (but im clearly confused :P)

Maybe there are still driver issues...

My temperature seems to be superb (~85 under full load), however I've had 1 black screen and then blue screen nvdvm.dll or so, after I started playing Windowed Mode and 7800 viewdistance, after about ~40 minutes playing.

Yesterday and the day before, I've been playing Crysis 1.2 full quality, and the new Alone in the Dark, for hours, and didn't experience any of these symptoms

In case the GTX 280, or at least it's early revisions, are flawed, I guess ATI HD 4870 or a more supercharged version of it, might be an alternative wink_o.gif However, down here in NL, only 512MB variants are available atm.

But I throw it on drivers for now.

Share this post


Link to post
Share on other sites

Don't give up yet. Try to test the card in another computer, go here http://www.nvidia.com/content....rl=http and download the system tools, try to monitor ur card temperature in load, maybe there's a bad contact between ur card and the cooler.

Share this post


Link to post
Share on other sites

@King Haggis:

I didnt mean that multi-rail cannot work with this card. I only meant what others brought forward before: In a multi-rail system you usually will have to manually balance the power consumption on the different rails, by looking up what your devices use, and dividing the devices over the rails.

Depending on the setup of the powersupply, this could be problematic in some situations.

That's for me basicly the reason to have only 1 rail, so you don't push the limits on 1 rail, while other rails are barely used, and the possible effects this can have.

@Temperatures

My own temps seem to be pretty awesome! (Updated them after using NVidia system software to monitor the temps)

Start Temperature: ~62

After ArmA Started: ~68

During Gameplay: ~75

In Forests, tested on Schmalfelden heavy veg: ~85

Been playing a while now, and haven't seen the blackscreen problem anymore. I remember when the blackscreen crash occured that I had Fraps running and tripple buffering enabled in Nvidia Control Panel, aswell as running ArmA in windowed mode.

ATM im running fullscreen, no fraps, and no tripple buff.

For some reason the performance atm seems a lot better.

I'm now trying to play with 7800m viewdistance, everything very high, incl AA, except terrain detail on High, as Very High option is not available at this viewdistance biggrin_o.gif

Share this post


Link to post
Share on other sites

Well, I'm back from my weekend of R&R and ready for another fight against my PC biggrin_o.gif

I downloaded Nvidia systemtools and started up Hellgate London, as that is the game that gives me most problems. When I was in the menu, the GTX280 fan became extremely loud as usual and temperatures were rising towards 90 degrees. I started the game and POOF! One minute later the black screen came on again.

I couldn't measure the temperature at that moment but after a restart, the GPU temperature was around 100 degrees so I assume it's a heat thing. I've ordered a replacement card. I hope this will solve my problems with the GTX280. I still want to enjoy the card, Nvidia has never given me any problems before. I'll try again with the replacement. If it fails I have to give the ATI a try.

@Sickboy, nice to hear there are fellow Nederlanders here :c)

Share this post


Link to post
Share on other sites

Last weekend I put in a friend's BFG GTX280 OC2 in my rig that has a Corsair HX520W and it worked fine, if that tells you anything. Got a 3dmark06 score of almost 18000, give or take 100 points. ArmA also ran great. I don't think your 650W was the problem  at all so I suggest you to get a refund for your 1000w unless you really want to keep it.

Share this post


Link to post
Share on other sites

I got my replacement GTX280 today and I am very happy to inform each of you who helped me that my troubles are finally gone (at least, for now).

My first GTX280 was a defect. The new one runs just great and hasn't given me any issues yet. It stays very cool too! Around 65 - 70 degrees when playing ArmA and 55 when idle. And the fan's not loud anymore. I'm so glad my system finally runs good again after all the trouble I've had with it. You should know I had several other issues before I posted here in this thread.

I'm keeping my new 1000W PSU. I've sold my 650W to a friend. After all, the PSU calculator told me I needed a 635W PSU with my "old" GX2 card so I needed a stronger PSU anyway.

I also managed to run Arma without the -dsound parameter. The reason I had to switch to Directsound was because a Soundmax driver for the onboard sound which came with my  Asus Striker Mainboard was still active. The driver was conflicting with OpenAl somehow. I have removed the Soundmax driver before I installed my new X-Fi card but somehow traces of it stayed. It really became a problem when Racedriver Grid wouldn't run, same story as with ArmA in my first post in this thread. The game was active in the taskmanager but didn't start. There's a big discussion going on in the Codemasters forums regarding this problem but the solution was to remove the traces of the Soundmax drivers. Somebody discovered this issue and explained how to remove the drivers manually by deleting a .dll file in the windows\system32 folder.

Maybe other ArmA players who used to work with their onboard Soundmax card (that came with their Asus P5 or Striker mainboards) but switched to a Creative card, have the same problems with ArmA and have to use the -dsound parameter now. The solution is rather simple, read this thread:

http://community.codemasters.com/forum....4180228

It should solve all OpenAl conflicts/issues.

Anyway, this topic has moved beyond the GTX280 now, I call it a day and want to say thanks once more to each one of you who showed interest and gave help when I needed it.  goodnight.gif   smile_o.gif

Share this post


Link to post
Share on other sites

Im very glad that you solved the problem. Now go play and have fun biggrin_o.gif

Share this post


Link to post
Share on other sites

Good to hear you fixed it. I am curious how your play experience is with that card. I myself just ordered an ATI 4870... I guess I'll be the guinea pig for anyone considering that card and playing arma. No one benches vid cards with Amra nener.gif

Share this post


Link to post
Share on other sites

Awesome nubbin; Let us know your findings smile_o.gif

Nice to Read King Haggis! Have you not had any blackscreen issues not even after prolonged play time?

After some more playing and testing, these are my results atm:

[*] Alone in the Dark and Crysis, I can play for hours at a time, without any problems

[*] When playing ArmA, usually after playing longer than 30 minutes, I can get black screen, and then a blue screen on nvidia driver, iirc: nvlddmkm.

[*] My GPU temperature during play is about 70-85, usually ~75.

I'm reading about other people experiencing this in other games, but usually within a few minutes. Example: http://www.nvnews.net/vbulletin/showthread.php?t=115245&page=2

As I was able to play Crysis, and Alone in the Dark, for hours, and it seems only to happen in ArmA after at least 30 minutes, I do not think I suffer from powersupply issues, nor do others seem to report bluescreens after their black screens, usually just complete turn off smile_o.gif

I'm guessing for now on Driver Issues, and possibly a combination of (forced) settings in the NV Panel and ArmA.

Unless anyone else has any ideas smile_o.gif

Will spend some more time on it soon smile_o.gif

update1

Hmm running prime95, errors after just a few minutes. Seems im having some heat problems with cpu. I'll update again soon.

update2

Cooler was full of dust and cpu-fan speed tuned too low. Prime is okay again, let's see how this inflicts ArmA, the big question is; how could I run aitd and crysis longer than ArmA :S.

Share this post


Link to post
Share on other sites

Well, I haven't had time to play long. I came home late yesterday evening. Besides, it's a little too warm a.t.m. in the Netherlands to spend much time behind the computer.

But I too got a blue screen in Armed Assault after playing for about 30 minutes and once when I shut down Windows Vista instead of shutting down I got a BSOD. Both times it had something to do with the videomemory. I believe it was the same BSOD that you (Sickboy) described. I'll take a look into this later so I can post the details of the BSOD.

The card stays cool while playing games and we both have good enough PSU's. I guess the BSOD's are a driverproblem. I hope Nvidia will sort this out in their next release. At least I can play ArmA for half an hour now instead of 2 minutes. Maybe there's another solution to the problem. I'll look into this later if the BSOD problem runs out of hand. Maybe we won't have these problems with Windows XP.

Share this post


Link to post
Share on other sites

After clocking the fsb a wee bit back, lowering cpu voltage by 0.04, and cleaning the cpu cooler,

the machine is Prime95 stable for at least a couple of hours, aswell as better temps.

I hope Nehalem is around the corner wink_o.gif

ArmA Results for now:

[*] Playtime: 160+ minutes, no black or blue screen yet

[*] Playmode: Ingame server+client, Coop, 1 player, there are some objects and vehicles+infantry moving around

[*] Island: Schmalfelden Heavy Vegetation

[*] Test situation: I'm running around and inside the forests Mid-West. At 50 Minutes, I switched to TroopMon, watching the ai's move around the Island.

[*] Viewdistance: 2500m

[*] Settings: Everyting very high, Terrain Detail High (Highest possible until 7800m VD), AA on normal, nvidia control panel aa-alpha set to multisample

[*] Resolution: 1920x1080@60hz Windowed

[*] Clocks: CPU 3265, MEM 980, FSB: 408. GPU: 670, GMEM: 2500, GShaders: 1460

[*] Software running: Everest to Monitor load, temperatures and voltages

[*] Temperatures: GPU: ~75-84, CPU SURFACE: ~37-43

My crashes before were on the same island, but probably with FRAPS running for FPS monitoring and as client on a deddy server, higher fsb thus also cpu & mem clocks and higher cpu temps.

Aswell as dust in CPU cooler and 400rpm lower fan-speeds (2x12cm on Cooler Master Geminii).

I guess there will be more certainty when i've been playing an evening of MP smile_o.gif

Also been playing Alone in the Dark for over 4.5 hours and no problems either. GPU Kept between 70-80 depending on scenery.

Share this post


Link to post
Share on other sites

I have that black srcreen followed by a restart or ctd also. But sometimes its after playing arma for 10 mins other 1 hour, and sometimes i can play for 2 hours without any problem (most of the times i dont have any problem). I also play with fraps on, and tried to reduce the cpu speed from 4.5ghz to 4.050 and i stil have the black screen sometimes.

Cant tell whats the problem.

Share this post


Link to post
Share on other sites

Me neither but I have to say I've always experienced CTD's and an occasional blue screen while playing ArmA, even with Windows XP and my 8800 Ultra, so I'm rather used to it. It's just that, with my old setup, I got these crashes and BSOD's maybe once or twice a week and yesterday, I had it twice on one evening. I hope it was just something random and not related to my new setup but I fear the worst.

But since this new hardware is still quite premature, I'm sure it's just a matter of releasing better Forceware drivers. If not, then we can only pray one day we'll have errorfree hard- and software that rids the world of the evil blue screens of death. That would be sooo nice inlove.gif (and no, I'm not talking consoles here).

Share this post


Link to post
Share on other sites
I hope Nehalem is around the corner wink_o.gif

While the Nehalem is very nice, ArmA doesn't support more than 1 physical core but it may be useful when OFP2 comes out. (It's supposed to multithreaded.)

I had a 5600+ before my Phenom and it was great, but I only upgraded my processor because I play a lot of Supreme Commander which utilizes all cores I believe.

Share this post


Link to post
Share on other sites
While the Nehalem is very nice, ArmA doesn't support more than 1 physical core but it may be useful when OFP2 comes out. (It's supposed to multithreaded.)

I had a 5600+ before my Phenom and it was great, but I only upgraded my processor because I play a lot of Supreme Commander which utilizes all cores I believe.

Luckily one uses the computer for more things than just ArmA.

Besides, why would 1 nehalem core, possibly overclocked, not be faster than a 1 kentsfield core?

I guess with OFP2 you mean ArmA2.

Share this post


Link to post
Share on other sites

I originally posted this on the other forums in response to a quite similar topic, I think it bears repeating here.

Quote[/b] ]Those are ambient temperatures I presume, it's a bit tricky to get run-time temps if it blows up while running. Considering those baselines though, it would not be surprising to see temps under load exceed the capacity of your specific chips.

Unfortunately, the dirty little secret of the electronics industry is that the operating environment temp ranges actually falls under what would be observed in icebox corp offices, not in random home settings.

Take for example the Dell 2408WFP monitor, I presume the specs are comparable to the 2407WFP of which I have two, and if memory serves me correct these numbers are actually slightly higher than the range listed for a CRT I had start smoking on me years ago.

Quote[/b] ]Temperature Range Operating:

5 ° to 35 °C (41 ° to 95 °F)

Temperature Range Non-operating:

Storage: 0 ° to 60 °C (32 ° to 140 °F)

Shipping: -20 ° to 60 °C (-4 ° to 140 °F)

Humidity Range Operating:

10 - 80% (non-condensing)

Humidity Range Non-operating:

Storage: 5% - 90% (non-condensing)

Shipping: 5% - 90% (non-condensing)

35/95 is a fair bit hotter than what I suppose may be typical for an air-conditioned house's ambient temps, but I want to point out a couple factors easily overlooked. Additionally, there are plenty of installations where for any number of reasons actual ambient temps easily approach the operating range maximums. They don't tell you about those ranges, and they're buried in the squinty-print, but they actually are a major concern.

1 - 35/95 is the upper limit of the operating environment, it's not unreasonable to expect that imperfect or deteriorated components will have a somewhat lower upper limit.

2 - General averaged ambient temps may be several degrees lower than ambient temps in proximity to the electronics, giving a higher than average temp supply.

3 - Component placement may in some cases artificially elevate immediate ambient temps, I suspect that a significant number of red-ring xboxen may have been shoved in closed cabinents under home theater systems, I've seen plenty in unventilated locations.

Unfortunately, what this all means is that you really need to be keeping the airspace around all your PC components well under 30/80 degrees, and preferably under the magic 72 degrees everyone swears by with aircon's. In addition, you need to keep all your thermally conductive surfaces squeaky clean and possibly re-grease any OEM greased heat sinks. Removing the cover really is a last ditch effort, and makes it get dirty a lot faster. You also have to make sure you have a way to get rid of that super-heated air.

Another factor that you need to consider is that multiple components, such as dual monitors, placed in close proximity, can easily boost the ambient temps between the devices to well outside the operating limits.

So what does this all mean? Well temps you see like those listed by the OP are generally observed at rest at the Windows Desktop. For power-saving and thermal reduction, a lot of the CPU and GPU is actually switched off at that point. The graphs in something like RivaTuner are especially effective in showing how once you 'switch on' DirectX 3D to create a render box, the power and temp loads jump 50% in some cases. Your CPU is 'resting' at 63/145, depending on the reporting method (socket vs on-chip) the actual temps may be a few to a lot of degrees hotter. You didn't indicate the CPU, so I picked at random an AMD Athlon X2 4800+. The numbers are going to vary, but the same point applies for other CPU's. Basicly, the operating range maximum is 70c, so your CPU could already be at its max if it's under-reporting, and you haven't even started MSN messenger yet. When you wake up the full CPU and start shoving data through it, it's going to jump well over warranty operating range limits, during normal operation. The same goes for your GPU and hard drives.

Now if you're bored with tons of money, a water cooled super rig, or even a mineral oil aquarium is still not going to solve your problem if you vent it all inside. You're going to need to duct and shroud a vent system to get that hot air out, and run a blown A/C in the side or wherever your intake fans are. I wish the industry would make more practical cases adequately utilizing 120mm fans, but all we get is 1337 cases with atrocious airflow routing.

We're all well acquainted with the false advertising rampant with PSU's (actual output down as low as 60% of rated on the label), I strongly suspect that actual thermal operating ranges are similarly suspect across the board.

Share this post


Link to post
Share on other sites
Quote[/b] ] ArmA Results for now:

Playtime: 160+ minutes, no black or blue screen yet

Playmode: Ingame server+client, Coop, 1 player, there are some objects and vehicles+infantry moving around

Island: Schmalfelden Heavy Vegetation

Test situation: I'm running around and inside the forests Mid-West. At 50 Minutes, I switched to TroopMon, watching the ai's move around the Island.

Viewdistance: 2500m

Settings: Everyting very high, Terrain Detail High (Highest possible until 7800m VD), AA on normal, nvidia control panel aa-alpha set to multisample

Resolution: 1920x1080@60hz Windowed

Clocks: CPU 3265, MEM 980, FSB: 408. GPU: 670, GMEM: 2500, GShaders: 1460

Software running: Everest to Monitor load, temperatures and voltages

Temperatures: GPU: ~75-84, CPU SURFACE: ~37-43

I wonder what your framerate was with those settings. I play the game @1680 x 1050 with everything set to very high, AA turned off and Post processing set to low, viewrange set to 2800 and get very decent gameplay but not something to write home about when compared with my old 8800 Ultra.

The 9800 GX2 I had sometimes produced an extremely smooth framerate followed by nasty stuttering and unreasonable framerate drops. The strange thing with that card was, sometimes gameplay was very smooth, another day the game would run like .... The new GTX280 at least gives a much more consistent framerate and the ArmA experience is very nice overall. I did expect a bit more to be honest but I still have everything set to default clockspeeds (Q9300 @2.5GHz. and GTX280 @ stock.)

I did a 3dMark 2006 run yesterday with AA turned off and with a resolution of 1680 x 1050 I scored a mediocre 12.500 points. My 8800 Ultra in Windows XP with a Q6600 CPU did the same. If I set the resolution to 1280 x 1024 I might get better scores but I would have expected around 15.000 with the 1680 x 1050 resolution. Maybe it's Vista, maybe it's normal.

I don't know my framerates in ArmA yet, visually it was quite good overall but not extremely smooth, maybe I'll measure them this weekend. It's good enough right now but I hope I can squeeze a few extra frames out of the game if I overclock things right. Yesterday, I played ArmA for 2 hours without a single BSOD or CTD. That's a good start  smile_o.gif

Share this post


Link to post
Share on other sites
Luckily one uses the computer for more things than just ArmA.

Besides, why would 1 nehalem core, possibly overclocked, not be faster than a 1 kentsfield core?

I guess with OFP2 you mean ArmA2.

True. You probably do a lot of compiling and encoding :O smile_o.gif

Plus the Nehalem processor will essentially be two Yorkfield quads slapped together on a single wafer, I read this somewhere. I'm a frequent visitor at Hardforum so it was probably there.

As of OPF2, I was reading a PC Gamer magazine that talked about Operation Flashpoint 2. The person over at Codemasters said that it's multithreaded so that even a low end Core 2 Duo would have both cores being utilized, therefore improving performance.

I don't know anything about ArmA 2, but it would be nice to have something like that in the next release! wink_o.gif

Share this post


Link to post
Share on other sites
Plus the Nehalem processor will essentially be two Yorkfield quads slapped together on a single wafer.

It will also be a more sophisticated design, featuring L3 cache and an AMD-style integrated memory controller amongst other things. I'll probably get one when the prices come down to replace my aging single-core Athlon 64.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this  

×