Nvidia's latest driver disables GTX 900M series GPU overclocks

[Syphadeus]

seems to me that this is the lesser of two evils, considering the bad press they would receive if people started damaging their laptops/themselves because of this 'bug'

Yes, however given the recent 970 debacle, Nvidia really doesn't need ANY evils right now, lesser or no. Again, this is an issue because as shown, Asus have actually marketed their systems as overclockable, which they now won't be. Consumers won't care whether the blame lies with Nvidia or Asus, but I would guess that anybody who wished to return their Asus notebook for a full refund using consumer protection on statement of goods and purpose would easily be able to do this and that isn't good business for Nvidia or Asus.

[GuidoLS]

There's absolutely nothing to keep people from down versioning their drivers, and there's seldom any real benefit to keep jumping up to the next new thing. As for Asus selling something as being overclockable, that's on Asus. Let them develop a driver that's capable. Or any other OEM, for that matter. Or do we want to go back to the good old days where processors of all stripes were locked?

[MrRockliffe]

I think overclocking isn't the issue, it's the ability to underclock. I can see how it'd be a massive benefit, being able to get the GPU to run at the same speed but lower voltage. Less heat and thus quicker GPU boost.

New drivers carry optimisations, that paired with better clock speeds would offer really great performance.

[jm101]

Another "blunder" that got thru quality control or a feature that someone put in?

[mercyground]

Just like their wonderful patch that prevents you from running physx on a card in a system that has an AMD card in it?

NV are idiots and this is a dick move. Your warrenty has always been void if you OC. This is just them being assholes.

[Corky34]

As for Asus selling something as being overclockable, that's on Asus.

Wouldn't they just say that you should be using the driver from their site.

[GuidoLS]

Wouldn't they just say that you should be using the driver from their site.

If they want to be the ones to guarantee/warranty the product, as well as shoulder the consequences when some or all of the laptop dies/melts/self-destructs, then yes - that's exactly what they should be saying and doing. But it doesn't happen that way, and we all know it.

Was it a d*ck move, as some are claiming? No more so than them cutting off all driver support to anything under the Fermi range of cards - that doesn't mean there aren't EASY ways around either problem. But if cutting out the ability to OC their mobile products is a company policy designed to save a lot of headache for both the user and company, then no, I don't believe it is. Plain and simple - Nvidia cannot and should not be held liable for what their OEM's do. Seriously - if Dell/HP/pick a vendor makes a crap computer, is it Intel/AMD's fault? Or MSI/Asus/Gigabyte makes a crappy MB? Because for a lot of years, the only way to OC CPU's from either company required some actual physical modification - sometimes to the MB, always to the chip.

You can't shoehorn a 350 big block into an old VW bug, wrap it around a tree and then blame VW when you do. (that's a 5.7l v8 engine for all you whippersnappers) Can't blame GM either, because in neither case, neither was being used as intended.

[DB007]

The interesting thing will be if the ASUS laptops (sold with the OC Ability) - will now be valid for refund if they can no longer OC.
Its never great relying on the manufacturer (e.g. ASUS/DELL etc) for driver updates as they often do not support the full-lifetime of the Nvidia driver updates that would be issued.

[ysluxus]

A better way may be that Nivida leaves options to disable or keep OC function. Overclocking laptop could be fun for users who want to do.

[magnusfl]

they should allow under clocking as that can save power but there right over clocking a lap top can be very damaging as there no way to add the extra cooling required to over clock

[Syphadeus]

If they want to be the ones to guarantee/warranty the product, as well as shoulder the consequences when some or all of the laptop dies/melts/self-destructs, then yes - that's exactly what they should be saying and doing. But it doesn't happen that way, and we all know it.

Was it a d*ck move, as some are claiming? No more so than them cutting off all driver support to anything under the Fermi range of cards - that doesn't mean there aren't EASY ways around either problem. But if cutting out the ability to OC their mobile products is a company policy designed to save a lot of headache for both the user and company, then no, I don't believe it is. Plain and simple - Nvidia cannot and should not be held liable for what their OEM's do. Seriously - if Dell/HP/pick a vendor makes a crap computer, is it Intel/AMD's fault? Or MSI/Asus/Gigabyte makes a crappy MB? Because for a lot of years, the only way to OC CPU's from either company required some actual physical modification - sometimes to the MB, always to the chip.

You can't shoehorn a 350 big block into an old VW bug, wrap it around a tree and then blame VW when you do. (that's a 5.7l v8 engine for all you whippersnappers) Can't blame GM either, because in neither case, neither was being used as intended.

I completely agree with the notion that you can't blame an apple tree for a bad apple pie if you put salt into it instead of sugar - but I think that your analogy is a slight oversimplification.

In order to fall back on the whole 'used as intended' argument, there must first be a clear statement of intended use set forth. The issue here appears to be the fact that this has changed somewhat between driver revisions. It is common industry practice to introduce new features with drivers. It is less common to remove features with new drivers. The issue here, unlike a car analogy, is common practice. Updating drivers on a notebook or laptop is a less common practice, but let us not forget that aside from features and optimisations, one of the main reasons for doing this would be fix bugs with certain games or programs at a driver-level.

When you consider this element, it's easy to say "Oh well, just use the old drivers then!" and that'll work perfectly fine until such a time that you run into a an issue with a game or program that is caused by a driver bug. Then you're faced with a decision to update the drivers to a version that fixes this, but in doing so you'll lose a feature of the product that could be important to you. If you're sold this feature of the product, why should the end user be faced with this decision and why should the 'you're not using it as intended' argument hold any water in such a case?