Nvidia back to clock-blocking 9**M.

Sources:

http://www.techpowerup.com/212769/nvidia-back-to-dirty-tricks-with-gtx-900m-series-overclocking.html
http://forum.notebookreview.com/thr...blocked-in-347-88-reblocked-in-350-12.771900/

I'm starting to kind have had enough of nvidias bs. Maybe it's about time to give the red team a go again.

 

I guess it's primarily the notebook manufacturers which don't want to deal with damaged notebooks because of some overclocks 'gone wrong'.

Implementing some switch in the BIOS seems actually reasonable to me.

 

It's about choice. It's always been an overclocker perrogative to do what he will with his hardare.

Also going back on their word just proves how untrustworthy Nvidia is.

 

Have fun.

AMD forced GPU manufacturers to voltage lock 7950 and 7970 models on a HW level which couldn't even be circumvented with a modified BIOS.

I had to send back one 7970 due to it and had to find an older unlocked model.

Funnily, I actually planned to undervolt the 7970 for most of the time (because it was used in a SLI configuration) which wasn't possible anymore, thanks to that stupid voltage lock, lol.

Btw, there is no such a thing like 'right to overlock'. If the hardware doesn't support it (laptop manufacture putting a voltage lock into the BIOS), complain to them and buy something else.

 

This is such bull**** reasoning. The OC limits and thermal throttling on mobile GPUs make it impossible to kill the GPU without tampering with voltage or removing the throttling measures through hacks.

The paltry +135 core clock limit for overclocking is nothing, and the only way you will run into problems is if the fans stop working and the throttling measures fail to work, and failure of those two things, the GPU would die even slightly underclocked.

 

I am eager to hear another good reasoning. Companies never spend any money (implementing that lock in BIOS and drivers) for no good reason.

If it's nothing, when what's the fuss about?

 

Protecting us from ourselves; sounds reasonable — NOT.

 

That's the question of the day. There is no competent reasoning to lock out a core clock increase of +135 with cards that are voltage locked and that have thermal throttling measures to cut back clocks when a certain threshold is released.

My personal beliefs is preventing OC'ing will push users to prematurely upgrade which gives Nvidia more money faster, but that can't be proven unfortunately.

EDIT: I am still able to OC on Windows 10|352.84 driver with an 860m. So it's just affecting 9xx series cards?

 

I think so. And only if there is a flag in the BIOS (put in by the notebook manufacturer I guess) as I understood.

 

Sure...as long as you don't expect them to also honor the warranty for running the product past its design envelope. You can't have it both ways.

 

Such an ignorant statement. The current OC limits prevent the "design envelope" from being pushed past. A +135 core bump is well within thermal limits under normal usage, and in abnormal usage you have thermal throttling which will scale back the clock anyways OC or not. Bumping the core by that much without modifying voltage or disabling throttling is completely safe.

 

No Nvidia, I was about to upgrade to GTX980m...

Either sort it out or I'm not going to bother.

 

I consider it my right to do what I want to with my stuff. I don't even know why you would have a different opinion, it doesn't give you anything. Unless you working for a hw manufactorer ofc.

Why. Overheating limits are in all modern hardware. As long as you stay within the heatlimits I don't see an issue here.

 

Why? I just stated why. Operating a product outside of its operating envelope is a valid reason for not honoring warranty. The same applies for almost every product you buy. You certainly can do whatever you want with your product as long as you don't expect the manufacturer to pay for it. I don't know why you'd have a differing opinion. I'm positive you'd feel the same if you produced a product and your customers wanted you to pay for them operating it outside of it's design. Nobody is saying you can't do that. The issue is you believing the manufacture should always cover you if you break it. Just how far do you extend it?

 

You meant to say an accurate and insightful comment and one that is the standard for nearly every product you buy. Overclocking any product is never completely safe. 'Such an ignorant statement'. :banana:

 

Voltage is not increased, and the heat threshold stays the same with a bump in core voltage. If the card can handle the bump without crashing there is literally 0 damage that can be done and would be considered completely safe.

Your comment was ignorant, which I pointed out, and it's funny you are trying to defend it still after I listed the reasons why it is so.

 

You are making a circular argument. I'm saying as long as you don't hurting the product then the warranty shouldn't be void. You are saying that when they are operating the product out of 'envelope' the warranty would be void. So since the operating envelope differs between different drivers, as long as they are running the correct driver they can overclock without voiding warranty. I agree.

 

To add to that, different driver sets can stress the cards at varying levels depending on if any improvements were made to certain applications which can cause the card to crash at previously usable clock speeds. So simply changing driversets, should in theory, void the warranty according to his argument.

 

Your comment is ignorant, which I pointed out, and it's funny you're still trying to defend it after I listed the reasons why it is so. Warranties are a fact. You're merely opining it's unlikely to damage your equipment. Yet it happens and you're arguing warranty should cover it. An ignorant stance.

 

You can't even defend your own arguments never mind trying to mistate mine. How ignorant.