When will we have a new driver from AMD? *drivers released today*

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

railven

Diamond Member
Mar 25, 2010
6,604
561
126
HDMI output?

Yes, in this situation it is an HDMI monitor, since my "every day" screen is HDMI or VGA only (stupid Samsung.)

I'd rather not have to swap my setup use, the gaming corner is fine, but since I'm usually sitting there after the PC is on, it's the every day spot that I experience it.
 

SickBeast

Lifer
Jul 21, 2000
14,377
19
81
Try 12.6whql and if you don't like those use 12.3whql. Also overclocking is ****ing yourself. I've had minimal issues with my 5870 since launch day, when I bought it. If you're not spending 10x as much time testing your overclocks as you are setting them- you're doing it wrong.

Set it back to stock unless you're wiling to put in the time necessary that overclocking requires (to do it responsibly, and not resulting in placing unfair blame on the card/drivers).
Only reason I say this is because I've been there myself. TEST TEST TEST TEST for hours and hours that overclock, now that is truly boring my friend- but it's the way it is if you want to OC.
What tests do you recommend? I've tried looping Heaven and Crysis. I'm not willing to do Furmark as that can kill the card.

I'm surprised that the card isn't Windows stable given the fact that it will loop Crysis for hours at 1220mhz.
 

Obsoleet

Platinum Member
Oct 2, 2007
2,181
1
0
What tests do you recommend? I've tried looping Heaven and Crysis. I'm not willing to do Furmark as that can kill the card.

I'm surprised that the card isn't Windows stable given the fact that it will loop Crysis for hours at 1220mhz.

I haven't overclocked for years but I did a far less scientific method when I OC'd GPUs. I would pick a general number (maybe slightly on the high end) that people were getting out of the same card online, then I'd use it- if I saw corruption or driver errors- I'd take it down a notch until I reached full stability.

The better way is probably to start at stock, and each week move the OC up 100mhz at a time on core/mem until you reach instability. I never wanted to do this, as it's more boring. First thing I'd do is find a definite stability point, then work from there. Too many variables between driver releases that might limit your OC, so I would stick with 1 solid release during the testing, that everyone pretty much validates as a good release for overclocking.

I'd start at my 12.2WHQLs that I have installed if I were to attempt an OC at this point. You'd have to start at 12.3WHQL I believe. I'm a hater on nonWHQLs but betas would work as long as they are found to be rock solid for your rig.

It's possible it's not your card but PSU or something else, but start at a stable point, control as many variables as you can, and go from there.
 

SickBeast

Lifer
Jul 21, 2000
14,377
19
81
I haven't overclocked for years but I did a far less scientific method when I OC'd GPUs. I would pick a general number (maybe slightly on the high end) that people were getting out of the same card online, then I'd use it- if I saw corruption or driver errors- I'd take it down a notch until I reached full stability.

The better way is probably to start at stock, and each week move the OC up 100mhz at a time on core/mem until you reach instability. I never wanted to do this, as it's more boring. First thing I'd do is find a definite stability point, then work from there. Too many variables between driver releases that might limit your OC, so I would stick with 1 solid release during the testing, that everyone pretty much validates as a good release for overclocking.

I'd start at my 12.2WHQLs that I have installed if I were to attempt an OC at this point. You'd have to start at 12.3WHQL I believe. I'm a hater on nonWHQLs but betas would work as long as they are found to be rock solid for your rig.

It's possible it's not your card but PSU or something else, but start at a stable point, control as many variables as you can, and go from there.
It's just hard to test and validate because the only thing that seems to kill the card is a wake from S3 sleep. I guess I could try sleeping it and waking it repeatedly. *shrug*
 

bononos

Diamond Member
Aug 21, 2011
3,889
158
106
Last edited:

DominionSeraph

Diamond Member
Jul 22, 2009
8,391
31
91
I really hope they will bring out a solid and stable driver within the next month or so. I'm tired of the crashes, hangs, and "display driver stopped responding" messages.

I know. I'm running 1720MHz on air and these darn drivers won't even let me get to the windows desktop!

Still waiting for Intel to get my 6GHz i5 2500k running right, too.
 
Last edited:

Obsoleet

Platinum Member
Oct 2, 2007
2,181
1
0
I know. I'm running 1720MHz on air and these darn drivers won't even let me get to the windows desktop!

Still waiting for Intel to get my 6GHz i5 2500k running right, too.

Ok, this was pretty hilarious. Quite a jackass move, as we all figured out what the culprit of instability is here.. but funny way to put it nonetheless :D

I read almost all of the "these drivers suck" threads, and it seems 90% of the time the issue is a conflict between some 3rd party OCing software and the drivers (and obviously the driver team has the right to do what they want, without worry about some 3rd party crap)- or simply an overclock that got thrown off stability due to changes in the driver that might push the card harder at stock clocks ect.
And of course there are legit bugs like the sleep issue on 7xxx.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
Well to be fair, some of you are being harsh on sick. The sleep issue with AMD drivers was very real, although I thought it was resolved a while back? Apparently not for everyone.

While i've liked AMD cards and have enjoyed the service of my 7970s when I had them -- I really hope AMD beefs up their software team, I feel like their hardware is great but they could use more staffing in the software area. Nvidia puts more money there and it generally shows. That is hard to deny.
 

djnsmith7

Platinum Member
Apr 13, 2004
2,612
1
0
Being an owner of a TriFire setup, I must say, it definitely seems like we're getting shafted on driver releases. I mean, for the $$$ we pay for these damn things, we should be treated like royalty, not like we picked em up at Ross.
 

BFG10K

Lifer
Aug 14, 2000
22,709
2,971
126
I would say most enthusiasts still don't do that, to stubborn in thinking that if its stable with the stress programs then it should be stable with everything else.
It goes much further than that. There’s a pervasive overclocking mentally on the internet where people actually think it’s a right to have overclocking.

I’ve seen people returning cards (“faulty”) because they couldn’t reach the same speeds little Timmy got on the internet. Others wouldn’t bat an eyelid at returning cards they’ve damaged through overclocking. Meanwhile, the rest of us have to pay for this RMA fraud.

Here’s the thing: if a card is rated at 900MHz and runs perfectly at 900MHz, but can’t do 901MHz, that’s still a 100% flawless card. Any “problems” are actually happening between the chair and the keyboard.
 

Final8ty

Golden Member
Jun 13, 2007
1,172
13
81
It goes much further than that. There’s a pervasive overclocking mentally on the internet where people actually think it’s a right to have overclocking.

I’ve seen people returning cards (“faulty”) because they couldn’t reach the same speeds little Timmy got on the internet. Others wouldn’t bat an eyelid at returning cards they’ve damaged through overclocking. Meanwhile, the rest of us have to pay for this RMA fraud.

Here’s the thing: if a card is rated at 900MHz and runs perfectly at 900MHz, but can’t do 901MHz, that’s still a 100% flawless card. Any “problems” are actually happening between the chair and the keyboard.

LOL Yep it happens.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
It goes much further than that. There’s a pervasive overclocking mentally on the internet where people actually think it’s a right to have overclocking.

I’ve seen people returning cards (“faulty”) because they couldn’t reach the same speeds little Timmy got on the internet. Others wouldn’t bat an eyelid at returning cards they’ve damaged through overclocking. Meanwhile, the rest of us have to pay for this RMA fraud.

Here’s the thing: if a card is rated at 900MHz and runs perfectly at 900MHz, but can’t do 901MHz, that’s still a 100% flawless card. Any “problems” are actually happening between the chair and the keyboard.

Oh, I couldnt agree more!
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
It goes much further than that. There’s a pervasive overclocking mentally on the internet where people actually think it’s a right to have overclocking.

I’ve seen people returning cards (“faulty”) because they couldn’t reach the same speeds little Timmy got on the internet. Others wouldn’t bat an eyelid at returning cards they’ve damaged through overclocking. Meanwhile, the rest of us have to pay for this RMA fraud.

Here’s the thing: if a card is rated at 900MHz and runs perfectly at 900MHz, but can’t do 901MHz, that’s still a 100% flawless card. Any “problems” are actually happening between the chair and the keyboard.

Hear, hear. I've seen discussions of RMA'ing fully working cards at other forums due to low overclocks and it is pretty ridiculous. Anyway, I nominate this as post of the year ;)
 

Schmide

Diamond Member
Mar 7, 2002
5,587
719
126
I'm using my 7770 for mostly streaming twitch.tv and own3d.tv and here are my findings.

12.4 was horrible, barely accelerated anything.
i then went to
12.7 beta better but the gpu usage was weird. It would jump high and low, causing the core clock to jump as well. I couldn't even play a LOL game while watching a LOL game.
12.6 is just way better. Flat lines. GPU doesn't jump to power save, etc. Now I can play a LOL game and watch a LOL game.