9600M GS Driver crashes from Overclock

So I have the m50VM-A1 and I've been overclocking the 9600M GS that it has the 600mhz core, 1520 mhz shader and 470 memory (stock is 430, 1075, 400)

The problem is, once in a while (maybe once every 3 days or so) the drivers will crash when I'm playing a game. It doesn't crash on any games in particular, so it seems random.

My temps never go above 70C and tends to stay around 65C.

There aren't any blue screens or anything, just the message that the driver has crashed and recovered successfully.

I'm using nTune for overclocking.

I've also ran 3mark and furmark, none gave artifacts or any errors.

So my question is, are my overclocks too high? Or is it driver related (I'm using the 195.62 drivers). BTW I've seen people post that they were able to clock to 650, 1600 and 500 and they didn't seem to have any problems.

 

The only way to say for sure if it's overclock-related or driver-related would be to either update the driver or try resetting the clocks to factory defaults and running for a week or two. If you still get those random crashes with the driver update, it's overclock related. If you still get them with factory default clocks, it's driver related.

 

Okay thanks for the response.

Another question is, would I be doing any damage to the GPU if I just restart my computer each time it happens?

 

You shouldn't be, no. If it were heat related, the computer would not be able to start back up until temperatures were below the critical level.

 

and just FYI, not every GPU is made equal, just because some others can OC to 650/1500/500 doesn't mean you'll be able to. This sounds to me like slightly too high of an OC. I'd drop either core, shader, or both by a bit.

 

I lowered it down to 580 core, 1490 shader, let's see if it crashes again.

 

Ive experienced similar in the past and it was the driver.

 

Its the same error I experience when I OC too high and try to game/run 3dmark. and I get it pretty often because my highest non-stable but passable 3dmark clock is 617/1100/1625. one out of 3 times mid-test I'll get that error(on 185.85 dox, 186.81, 186.82, 195.62) but won't get it at 610/1000/1550.