EVGA Precision

[falcon409]

I asked this in another thread but it appears it will be ignored, so I'll start my own, lol.

I read all the time in here about folks adjusting fan speeds and adjusting GPU this and OC that, but what is really possible and far can you go before you melt your system into a little ball of molten lava?

I have downloaded the latest version of EVGA Precision. Nice interface, pretty straight forward settings, but in the hands of a novice at this, any one of those could be a killer to my system. So how do you know what you can and cannot achieve beyond the preset limits of your card and fans?

I run an NVidia 9600GT with 512 ram and 8gig memory.

[harleyman]

The read me states to go in very small incriments for overclocking both core clocks and mem clocks...

Then test ...If it fails you will know..

As far as fans, cank them to 100% and see your temps drop..then back off to 5060% or what suits your system and ears...

[DB93]

Hi Falcon,

I've personally had no luck with trying to run EVGA Precision's Test mode under Vista64, though it worked fine for me under XP32, so your mileage may vary for using the test function. Other than that, the best I can suggest is to make sure that the checkbox for 'Apply Settings at Windows Startup' (or whatever the exact wording is, I'm not at the machine that I run precision on at the moment) is NOT checked while you're testing. That way, whenever you reboot, the card will go back to its default settings.

This is helpful should you crash out or lock up while playing with the settings, as those settings won't be re-applied upon rebooting. Once you've got settings that you're happy with (and of course, are stable), then you can enable the 'Apply Settings on Windows Restart' option to have those settings enabled "permenantly" (I say that in quotes, because you can always re-run precision and disable the restart option, or change any of your settings.)

Remember to keep an eye on the card's temp, as well as keeping a lookout for strange artifacting on-screen while playing with your settings and testing, as those will be the first indicators of pushing the card too hard. Also, as harleyman said, go in very small increments and just keep playing until you're happy with the results and the card is stable. Personally, I'm running an EVGA 9800 GT 1GB card, and anything over stock card speeds cause bad artifacting and instability for me, however I still use the utility to crank up my video card's fan to 100% before any extended FSX time. :)

Good Luck!

[Bjoern]

Raise core clock by 10 Mhz, run 3D Mark, raise memory speed by 10 Mhz, run 3D Mark, repeat till 3D Mark freezes, kick back the clocks a notch until they run stable, voilà, GPU and VRam overclocked.

P.S: The shader clocks start throwing errors first, so once you run into that wall, ease up on the shader clocks a bit, unlink it from the GPU clock and go on overclocking just GPU and memory.

P.P.S: Leave the fan speed on "Auto". A card running louder than necessary isn't quite pleasant on the ears.

[kilo delta]

I wouldn't bother to overclock the GFX card at all,tbh.......the gains are negligible. It'd be no harm to increase the fan speed,though, especially if you're planning on extended periods of gameplay.

[falcon409]

According to this, with one GPU fan attached to the card and one additional fan running 100% the GPU temp is now "down" to 39C, it was 44C initially.

[rwmarth]

So related to this, is it possible to break your otherwise fine running card by keeping the fan speed up? Certainly if the card has issues before it could hurt it by cranking it up to 100% constantly I would imagine, but are the cards designed to handle it 100% for long periods of time?

[falcon409]

So related to this, is it possible to break your otherwise fine running card by keeping the fan speed up? Certainly if the card has issues before it could hurt it by cranking it up to 100% constantly I would imagine, but are the cards designed to handle it 100% for long periods of time?
Apparently I need a better fan setup. My GPU temp never drops below 37C (98F) and in the middle of an FSX flight this morning it reached 49C (120F).
That's with two fans on the card and two on the CPU. It would be a lot higher if I hadn't added an additional fan on the GPU. I don't know how folks keep all these low temps, they must have more than just fans to cool their systems.

[Dangerousdave26]

Remember that cooling is all relative to the ambient air temperature in the case, which is being heated by all the other devices. The temperature in the case is dependant on the ambient air temperature in the room that the whole PC is in.

So some times what you need to do is provide more cooling directly into the case.

Outhouse member Great Ozzie (http://www.sim-outhouse.com/sohforums/member.php?u=80376) has a reasonably good solution for this problem.

You might try duplicating it. It looks fairly inexpensive.

http://www.sim-outhouse.com/sohforums/showpost.php?p=295809&postcount=15

:icon_lol:

OK Ed that was all tongue in cheek but cooling is directly affected by the air temp going into the PC. If there is a problem with that then all other fans in side will have a hard time keeping up.

If you have room in your case try adding another case fan to push the cooler air in and just as important take the air out.

Those fans are not usually that expensive and in most cases can not hurt.

[falcon409]

Thanks, I'll take a look at what else is out there and set to work getting a better cooling system.:ernae:

[rwmarth]

That's with two fans on the card and two on the CPU. It would be a lot higher if I hadn't added an additional fan on the GPU. I don't know how folks keep all these low temps, they must have more than just fans to cool their systems.


As nice as cool temps are, I think people can get too carried away cant they? I mean 50 degrees isn't that hot for a graphics card under load. They certainly are designed to run hotter than that even. But of course heat isn't exactly a good thing either! Im more afraid of doing harm to the card by running the fan at high speeds constantly than temps around 50. And so I for one would love to know a bit more about this all....

[harleyman]

So related to this, is it possible to break your otherwise fine running card by keeping the fan speed up? Certainly if the card has issues before it could hurt it by cranking it up to 100% constantly I would imagine, but are the cards designed to handle it 100% for long periods of time?



The fans can and will run 100% all the time..They are designed to spin that fast and will.

Most cards never need 100%...My 4870X2 has been running at 80% for 1 1/2 years now... It also has a lifetime warrenty....

[Bjoern]

I never hear my GTX275, even during benchmarks.


Then again, it has to battle the noise level of six 80mm fans first anyways.

[falcon409]

So, I assume it's safe then (since no one made any mention of this) that my GPU running at 120F during FSX sessions is nothing to be concerned about?

[Bjoern]

So, I assume it's safe then (since no one made any mention of this) that my GPU running at 120F during FSX sessions is nothing to be concerned about?

Depends on how much 120°F are for us Europeans. ;)

Sounds like something around 50°C, which is totally okay for a video card. You should start worrying at 70 to 80°C.

[falcon409]

Ok then, that works for me, Thanks!:jump:

[rwmarth]

The fans can and will run 100% all the time..They are designed to spin that fast and will.

Most cards never need 100%...My 4870X2 has been running at 80% for 1 1/2 years now... It also has a lifetime warrenty....


Good to know. Thanks!

[kilo delta]

Perfectly acceptable temps....my old 9800GX2 would run in the 80°C 's!

[falcon409]

Perfectly acceptable temps....my old 9800GX2 would run in the 80°C 's!
Whoa nellie, lol. Guess mine is fine then.:applause:

[GaryGB]

Hi All:

Remember: most PC hardware manufacturers state that for every 10 degrees over their designated operating temperature range that a device is run at, the life of that device may be effectively cut in 1/2. :engel016:

Also, be aware that manufacturers may contest and deny a warranty claim if they are able to determine that you overclocked a device... even if one used the manufacturer's own software to do the overclock !


That being said, I do also overclock my own system components, but I now try to stay safely within the manufacturers designated operating temperature range (but when overclocking regardless of the rated range, never with any more than 10 degrees temperature increase above the "rated" range during a "high demand" session in FSX).


Personally, I would always want to first run an 'accurate' monitoring utility so the GPU temperature is displayed on-screen during FSX flight time in both windowed and full screen 3-D mode, to see what ones maximum rendering engine impact on temp is going to be, then limit the overclock accordingly.


Hope this info might help in deciding for vs. against, and if so... "how far to go" when contemplating an overclocking scenario. :mixedsmi:

GaryGB

[hey_moe]

My two cent....I OC the processor and memory, but have never seen that much of a difference OCing the Vcard. Have any of you guys seen that much of a difference...Mike