Mal X said:

i'm going to ask a couple of dumb questions:-

how do i check my cards temp whilst gaming, or is it possible to check what it was after i quit gaming

plus, should the cards fans be on auto, or manual/ custom, because right now they're on auto 30%

thanks

Click to expand...

As the other chap said, look in the afterburner config, check GPU temp and tell it to display in OSD and you're ready to roll

As for fan control its up to the user, depending on what card you have you may get away with leaving it auto, the Lightnings TFIV cooler kept things nice and cool for me on auto.

BrianB said:

What's the max safe volts on air for a typical 7970? Can't find the answer on google. Max temp 54c in heaven at 1200/1600 1250mV.

Click to expand...

For a ref 7970 I'd say 1.25v is getting pretty close to the limits I personally would be happy with on air, but each to their own, one thing to keep in mind is the temps of the VRM's, not all cards show a temp read out for them, they can get a lot hotter than the core.

Have you hit 1400 core yet?

Not yet and without 1.4v on the core I don't think it will be possible unfortunately.

@pgi, glad to see it going well mate, without me trawling through this thread, what was your max on air??

The max I was comfortable with on air for benching was 1290/1750 @ 1.287/1.65v, for gaming use I had no reason at the moment to push the clocks above the stock 1070/1400.

Cracked a window open first thing this morning to get my ambient temps down, currently idling at around 16c

Just going to tune up my CPU and memory then back to work on the GPU.

Patching single link to 120hz!!

Firstly we need to thank Vega for finding this and ToastyX from 120hz.net/monitortests.com for creating this patch.

Original guide: http://www.monitortests.com/forum/Thread-AMD-ATI-Pixel-Clock-Patcher

All I'm going to do is throw in some fool proof screens of what to do, this is now a hell of a lot simpler than it was a few months ago too

1. Download atikmdag-patcher-test.zip from above link (if this doesn't work let me know and I will mirror it). Inside will be 2 .exe files, atikmdag-patcher.exe and testmode.exe

2. Open up atikmdag-patcher.exe, it will say (as in pic below) what limits it has found and if you want to patch them, press 'YES'. Once patched you can close atikmdag.


3. Open Testmode.exe, you will simply get a small window asking you to enable test mode, click yes, once complete exit.

4. Download Custom Resolution Utility from the above link and open CRU.exe.
Click the 'detailed resolution' highlighted with blue with a black arrow pointing towards it, then click 'Edit' which the red arrow is pointing at:


5. In the Edit box, the area I have boxed in red, it will probably say "60.00hz" Change this as I have to "120.00" Press OK and exit CRU. Now reboot your machine.


6. Open up your advanced monitor options and you will now be able to force refresh rates up to 120hz


Hope this helps those that need it, to push 120hz over a single link DVI connection

Wow it worked!

I couldn't quite get it up to 120hz, I get a fuzzy screen. But I got to 100hz which for me is more than enough.

Thank you very much!

Glad it worked to some degree for you, it may be worth trying a better DVI cable if you want that extra 20hz.

Edit: Also as it stands I have the single fastest (score wise) 7970 on the 3dmark11 thread clocking in at 11653 graphics score with clocks of 1345/1970, I'm pretty sure I've hit a wall on the core, 1350 or over gives me driver crashes, the memory looks like it may even push 2000mhz

Working on heaven tonight, I'm not expecting miracles from the score however as I can't get close enough to the core speed of the top 7970's as they all use over 1350 core, provided my memory holds up I'll be running at 1340/2000, failing that 1340/1975.

Probably selling mine very shortly as cant really justify keeping it now Iv'e got another card. Shame really as its only seen about a couple of hours day light.

For a ref 7970 I'd say 1.25v is getting pretty close to the limits I personally would be happy with on air, but each to their own, one thing to keep in mind is the temps of the VRM's, not all cards show a temp read out for them, they can get a lot hotter than the core.

Click to expand...

Thanks! I have mine at 1262 at the moment for 1300/1600 seems stable and also temps rarely reach 60c with my new cooler. The VRM is cooled stock and now also has fans on it, also all the components I could mount heatsinks to now have them. No idea how to get the VRM temps for my card, maybe I can't gonna google it.

Look's like that, except for the VRM heatsink is a pre installed one.

Sorry, I know it's the wrong thread!

Ooo I found the VRM temps in GPU-Z gna try out in game.

Max VRM1 - 84C averages 69C VRM2 74C averages 62C, think it's okay?

I can hear some coil whine/buzz though. That normal?

can any lightning owners run this please and report back what you get its a memtest for open cl cards http://folding.stanford.edu/English/DownloadUtils

direct http://fah-web.stanford.edu/MemtestCL/memtestCL-1.00-windows.zip

my results


note random blocks, ran this due to boot error code 62 which points to gpu


may have to rma my card

Coil whine is normal on the 7970's, I have a little on my lightning, but I only noticed it once the cooler was removed.

Geeman did you clock the card any? Going to be after a second card soon

Nickolp, how long does the test take? If I have time I'll give it a run tonight.

pgi947 said:

Coil whine is normal on the 7970's, I have a little on my lightning, but I only noticed it once the cooler was removed.

Geeman did you clock the card any? Going to be after a second card soon

Nickolp, how long does the test take? If I have time I'll give it a run tonight.

Click to expand...

about 5 mins max, odd though just ran again and no errors, puzzled now

I only notice it when my ear is close to the pc so I guess it's not that bad. The vrm temps are about 20c lower when voltage is 1068 lol. Duno what I am going to do.

Hmmmm, max 57-54C on my VRM when not overclocked, and I can't tell the difference in the games I play. 7970 is too beastly barely needs an overclock.

Nickolp1974 said:

about 5 mins max, odd though just ran again and no errors, puzzled now

Click to expand...

I'll give it a couple of runs in a couple of hours and report back.

BrianB said:

I only notice it when my ear is close to the pc so I guess it's not that bad. The vrm temps are about 20c lower when voltage is 1068 lol. Duno what I am going to do.

Hmmmm, max 57-54C on my VRM when not overclocked, and I can't tell the difference in the games I play. 7970 is too beastly barely needs an overclock.

Click to expand...

To put it into perspective, my VRM's with 1.365v pushing through peak at around 45c under water, on air at 1.287v they would sit in the mid 60's with a custom fan profile. Hotter VRM's are less efficient, nobody is sure on the temp limits of them either, MSI told me 'the cooler the better' so make of that what you will.

The cooler the better?! LOL you are blessed to have had this wisdom shared with you! Is it the heat or the voltage that kills cards? 84C sounds okay to me for VRM, I think they are rated at 115 but some are rated at 85C. 84C seems to push 1 of those limits.

I have no idea what they are rated at, IMO voltage is far more likely to kill a card than heat, that said increased voltages can raise temps which in turn can cause issues. Just take it easy and keep an eye on the temps and you'll be fine

Thats about all I could get out of MSI when I asked how to increase voltage limits and how to get hold of Afterburner Extreme, the rest is very cloak and dagger, with only chosen overclockers given the privilege of the modified BIOS files to enable over-voltage up to 1.8v.

Nickolp1974 said:

can any lightning owners run this please and report back what you get its a memtest for open cl cards http://folding.stanford.edu/English/DownloadUtils

direct http://fah-web.stanford.edu/MemtestCL/memtestCL-1.00-windows.zip

my results

note random blocks, ran this due to boot error code 62 which points to gpu


may have to rma my card

Click to expand...

I had an error 62 code on my first card after about 2 months usage, I RMA'd it and got a new one, it took around 3 weeks from the day I sent my card.

StonerDude said:

I had an error 62 code on my first card after about 2 months usage, I RMA'd it and got a new one, it took around 3 weeks from the day I sent my card.

Click to expand...

is it ok now then??? did it sort your problem???