NVidia Puts the Kibosh On Overclocking of GTX 900M Series 138
An anonymous reader writes Nvidia surprised members of the overclocking community this week when it pulled OC support from drivers for its 900M series mobile graphics cards. Many users (particularly those who bought laptops with higher-end cards like the 980m) were overclocking – until the latest driver update. Now, Nvidia is telling customers not to expect OC capabilities to return. “Unfortunately GeForce Notebooks were not designed to support overclocking,” wrote Nvidia’s Manuel Guzman. “Overclocking is by no means a trivial feature, and depends on thoughtful design of thermal, electrical, and other considerations. By overclocking a notebook, a user risks serious damage to the system that could result in non-functional systems, reduced notebook life, or many other effects.”
Comment removed (Score:4, Insightful)
Re:"risks serious damage to the system" (Score:5, Insightful)
because quite often it takes a lot of effort to identify that the cause was user overclocking, by then the customer has complained to the store, many stores have policies of replacement or money back in first X days. While it definitely should be a try it at your own risk situation, the reality is people will basically lie to the retailers face saying they did nothing and expect a refund/replacement.
Bollocks. (Score:4, Insightful)
NVidia are merely doing what Intel did with clock locking. And for the same reason: they don't want people getting a cheaper card and overclocking it, they want people to buy the most expensive card.
OC doesn't lead to catastrophic failure in the first year unless the system already has a fault that would appear in a few years of normal use. This is how they test MTBF: stress test and see failures, scale back to the rate of normal use to get the mean time before failure.
However, that means that without OCing, the system would last on average longer than the manufacturers' warranty,which they scale so that they don't have many failing during that time. Add in that most won't have the receipts more than a couple of months and their fear is that OCing would get some warranty returns within the period they have kept the receipts for.
Re: (Score:2, Insightful)
You can't beef up the cooling in a laptop, there's a big difference.
Re: (Score:3)
No, they test the system with the rated TDP and ship it, and IF they do any lifetime testing at all (which is very dubious, depending on the vendor) they do what another poster up the chain described and intentionally cook the device and make sure the mean time to failure fits in a window their quality guys establish for the rated TDP, not the overclocked point. If it were able to work beyond the rated TDP for the warranty period, that means there's some material they could have removed, but didn't. If you'
Re:Bollocks. (Score:4, Insightful)
Chalk this decision from nVidia as a few assholes ruining it for the rest of us.
Re:Bollocks. (Score:4, Informative)
The real problem with OCing and failure, is that when a failure does occur sometimes you can't tell it has occurred. Part of the chip just stops working that might not be critical to the system. It's near impossible to tell this has happened without a lot of analysis that the average person overclocking is simply not capable of. Just because it happens to "work" when you first do it does not mean you aren't causing irreparable damage.
Mobile (Score:1)
Except that these are an "M" (for Mobile) series, not a card, but a chip. Also not something you buy individually, but rather what comes with the laptop, and makes the entire thing useless if it burns out.
For desktops, you can add cooling to support the OC, and if you do burn out the card it can be replaced. Few laptops support hardware upgrades or replacements for GPU.
Re: (Score:2)
Re: (Score:2)
Desktop cards are easy to replace and you're only out the cost of the card. Fry the GPU on the laptop and in a lot of cases you have a repair that costs as much as the laptop did new. NVidia doesn't want to anger their customers, but there is probably enough hardware that they're paying warranties on that this has become necessary.
Re: (Score:2, Troll)
That's fraud and they should be charged with such. A few cases of fraud going through and that shit will stop real fast.
Re:"risks serious damage to the system" (Score:4, Insightful)
ROFLMAO at your Naivete.
People are copying our software!
That's piracy and they should be charged with such. A few cases of piracy going through and that shit will stop real fast.
People are driving drunk.
That's drunk driving and they should be charged with such. A few cases of drunk driving going through and that shit will stop real fast.
People are murdering other people!
That's murder and they should be charged with such. A few cases of murder going through and that shit will stop real fast.
Nowhere in the entire history of mankind has a few people being punished for a crime stopped others from committing the same crime.
But hey, you're an optimist so you've got that going for you. And that's nice.
Re:"risks serious damage to the system" (Score:4, Insightful)
Punishment for a crime is just ritualistic sadism.
It takes the offender out of normal society and and puts them into a society full of criminals. It gives them a record to make it more difficult for them to integrate into normal society. it makes the best possible effort to ensure that they reoffend.
The only reason for locking someone away is if they are a danger to society. Since most people locked away in the US are not a danger to society, we have a nation of sadists. And, as the prison system is privatised, an industry of businesses profiting from sadism.
Re:"risks serious damage to the system" (Score:5, Insightful)
Re: (Score:1)
Re: (Score:3)
The only reason for locking someone away is if they are a danger to society.
That's a good reason but it's not the only reason. Punishment also acts as a deterrent to those who would commit crime and helps to balance the scales of justice by ensuring that there are consequences for bad action. Currently we use imprisonment way too frequently and the system is poorly designed but that doesn't mean there is no appropriate use for it.
Re: (Score:3)
A wonderful theory upon which our penal system is based. The only problem is that extensive research suggests that it doesn't actually work that way. The vast majority of individuals who commit crimes operate on the assumption that they won't get caught, and thus the potential consequences are largely irrelevant to them. And don't even get me started on so-called "white-collar crimes, which are as a rule *far* more damaging, and in practice carry few if any penalties.
Re: (Score:2)
The vast majority of individuals who commit crimes operate on the assumption that they won't get caught, and thus the potential consequences are largely irrelevant to them.
So you're choosing as a sample the people who weren't deterred and using that for evidence that deterrence doesn't work?
Re:"risks serious damage to the system" (Score:4, Insightful)
Nowhere in the entire history of mankind has a few people being punished for a crime stopped others from committing the same crime.
Fear of punishment doesn't have to stop everyone from committing a crime. It only as to stop as many as needed to keep the numbers down to something people can live with.
Re: (Score:2)
because quite often it takes a lot of effort to identify that the cause was user overclocking,
not really. It's pretty easy to design the chip with a fuse or other equivalent feature that will permit finding out rapidly whether the user was overclocking and/or overvolting, and if the GPU overheated. the unit gets sent into the depot, jump pins are used to connect to the GPU and find out if the fuse is blown, and the warranty is either approved and the machine is replaced and the fact that the GPU failed without overclocking, overvolting, or overheating. of course, thermal control is typically the res
Re: (Score:3)
And of course it is impossible for the fuse to be blown by anything other than the user overclocking, right?
Re: (Score:2)
And of course it is impossible for the fuse to be blown by anything other than the user overclocking, right?
It's not impossible, but it's unlikely to happen by accident.
Re: (Score:3, Insightful)
Then it is completely worthless. Instead of having to try to determine whether the problem was caused by overclocking, you instead get to try to determine why the fuse was blown. Unless, of course, you are planning on them just rejecting warranty claims because it is 'unlikely' that it is their fault. OIn which case they might as well not include such a charade and just claim that it is 'unlikely' any problem is their fault and reject all warranty claims.
Re: (Score:2)
Re: (Score:2)
My point is that EVERY legitimate warranty claim is the result of something happening that should not have happened. So here we have a proposal to deny someone's warranty claim based on the fact that an eFUSE or whatever said they overclocked the chip, when in fact the state of the eFUSE itself could be a defect. That makes no sense at all.
Re: (Score:2)
Don't think I'm siding with NVidia on this, though; I'm ju
Re: (Score:1)
It isn't about "a chip". It's about a system that is designed for a specific thermal and electrical load. nvidia probably got flak from notebook makers who were facing dissatisfied customers.
You only have to look at a lot of the nonsense comments throughout, such as yours -- people just contriving how "easy" everything is, and how simple it is. Yeah, and I'll bet all of you design notebooks. No? Then shut up.
Re: (Score:2)
Re: (Score:1)
That's not how normal consumers think, and protecting yourself from bad PR for failing hardware is important. It doesn't matter if it was user error, your company name is attached to the negativity. Look at the toyota call back for model that was supposedly accelerating by itself, turns out it probably wasn't but toyota still took a PR hit for the episode. https://en.wikipedia.org/wiki/2009%E2%80%9311_Toyota_vehicle_recalls#Driver_error
Re:"risks serious damage to the system" (Score:5, Insightful)
Because NVidia got seriously hammered not that many years ago by 'Bumpgate' when their laptop GPUs were having serious reliability problems with their physical connection to the circuit boards, mainly caused by heat.,
While people like to claim of course they did nothing wrong, I am sure people who cook their laptops overclocking them will always try and point the finger back at NVidia...
Hence, they are playing it safe.
Desktops have MUCH better cooling systems, and hence are much less likely to suffer from extreme temperature problems...
I suspect it is also a sign they are pushing the limits harder - remember, new generation GPUs have built in 'overclocking' in the form of dynamic clocks already,
so they are using up the headroom they had more effectively. This means you are more likely to be pushing past a limit, and less likely to notice (until too late).
They will always wear the fallout from such peoples actions.. so they have obviously decided right now the risk is not worth the reward.
Re: (Score:3)
Re: (Score:3)
Re: (Score:3)
Comment removed (Score:5, Insightful)
Re: (Score:2)
I just blow a few doobies and maybe shotgun a beer. Dust will eat your brain, man.
Re:"risks serious damage to the system" (Score:5, Informative)
Because NVidia got seriously hammered not that many years ago by 'Bumpgate' when their laptop GPUs were having serious reliability problems with their physical connection to the circuit boards, mainly caused by heat.,
While people like to claim of course they did nothing wrong, I am sure people who cook their laptops overclocking them will always try and point the finger back at NVidia...
The Quadro FX1500 and some other chips had a die bonding problem. nVidia failed and they knew it, and what's more, the downstream OEMs knew it. I had an HP Elitebook with FX1500 graphics, which died the death of a thousand dogs amen. It took me literally over 24 hours of phone time to get a replacement because fuck you, HP, that's a name for a sauce, not a computer company. But what really clinched the deal was that I found out during the course of the problem that this was actually a known problem, techs inside HP (the high-end ones, not just phone monkeys) actually knew of the problem. And this was a laptop with an MXM video card, so they could have replaced them. But since this is HP, they didn't do that. And no manufacturer will, which is why paying extra for your notebook's video card to have an MXM connector is a boondoggle. In theory, they can replace just the video card. In practice, that costs too much, they will just replace the whole machine if you can get them to admit that they made you a lemon.
So yes, nVidia just fucked up. No, I never OC'd my mobile Quadro. Yes, that's the part of the PC that died, before the HP "tech" came out and killed it the rest of the way (it would work until it heated up, before that) with static, days after I called. That's what your on-site three-year warranty from HP gets you. Fucked.
Re: (Score:1)
Because NVidia got seriously hammered not that many years ago by 'Bumpgate' when their laptop GPUs were having serious reliability problems with their physical connection to the circuit boards, mainly caused by heat.,
They didn't get hammered.
Especially when you consider what Nvidia did. They knew it was bad, lied to investors, manufacturers and the public, and then, to appease them, they offered them a garbage Compaq laptop.
If you were even that lucky. Because Nvidia left a lot of the repairs and compensation up to the manufacturers, many of us got completely screwed.Those of us with high end Sony laptops were left high and dry. Not that I would have accepted their peace offering. Even today, 6(?)years later, my Sony is
Re: (Score:3, Insightful)
You've never worked customer service before.
Rule #1 - The customer always lies.
Rule #2 - The customer support person is not a technician.
What likely happened is that some laptop brand (let's just point at Alienware) was getting a high return rate, and the GPU was found to be the culprit, and a couple of idiots left the OC settings on the hard drive when they returned it, and someone took notice.
The subnotebook/tablet/ultrabook design can literately not afford to be overclocked since the design of the laptop
Re: (Score:2, Informative)
Customer support always lie.
Customer support don't ever want to support the customer.
Customer support is not a technician.
Customer support has a script for non-technician and will insist on going through every bullet point.
Customer support is not there.
Re: (Score:3)
Because it may not be the end user who has done the overclocking. The supplier may have in order to claim a higher spec to the hardware than NVidia is willing to support.
Also, by providing the option it could be claimed that NVidia is supporting the option. That makes them liable if it causes a problem, particularly something nasty like overheating and an electrical fire.
Re: (Score:3)
If a woman puts her wet dog in the microwave oven to dry, who is responsible?
Now compare that to how a US judge what answer those questions.
I rest my case.
Re: "risks serious damage to the system" (Score:5, Interesting)
Re: (Score:3)
If the user overclocks their GPU and it ends up overheating and breaking down isn't the responsibility for that on the user's shoulders? Why does NVidia care so much?
Normally, yes. But if the driver has explicit support for it, not so much. Basically nVidia is washing their hands of legal responsibility for you breaking your laptop.
Re: (Score:2)
Does the law somewhere state that NVidia is still responsible for the damages since their drivers have such an option or what is missing from this story?
In many countries the manufacturer can't absolve themselves from responsibility if a user accessible feature is capable of causing hardware to self destruct. e.g. Australia where people have successfully claimed warranty repair on botched aftermarket firmware flashing of their phones, Apple has been forced to change advertising of extended warranties time and time again, and Microsoft has been forced to honour warranty replacements on 4 year old Xbox360s which red-ringged despite only offering a 1-year warr
Re: (Score:2)
Re: (Score:2)
I'm trying to figure out how a user could possibly *be* responsible when it comes to overclocking something in a notebook.
With a desktop, you've got a lot of variables that aren't just within the end-user's ability to assess, but actually entirely within their ability to assess and control. You've got thermal readings, understanding of fan speed, placement of cables and other things that affect airflow management, and the ability to choose just how much (and what form of) cooling is in the case to begin wi
Because of support cost (Score:2)
Re: (Score:3)
If the user overclocks their GPU and it ends up overheating and breaking down isn't the responsibility for that on the user's shoulders? Why does NVidia care so much? Does the law somewhere state that NVidia is still responsible for the damages since their drivers have such an option or what is missing from this story? If some law somewhere did state that then I could totally understand NVidia's stance, but at the same time it would make me wonder why it doesn't apply to desktops, then.
if Nvidia provided the capability and it caused the damage they might wind up liable for the damage, wether from a suit or a manufacturer making them cover repairs. To use a car example, if an engine control module lets you fiddle with settings to get more horsepower or remove the rev limiter and you lunch an engine after using a form the manufacturer capability in the system then they may be liable for repairs.
Re: (Score:2)
It makes sense that the driver package enabled overclocking should be a feature the manufacturer has control over. It's senseless and stupid to overclock a notebook GPU in the first place.
nVidia probably started getting calls from one of the OEMs like Dell or HP showing that many expensive warranty replacements were tracked
Re: (Score:2)
If the user overclocks their GPU and it results in a fire, what happens then? I suspect the motive here isn't just about preventing users from shooting themselves in the foot. It's about preventing them from melting it off in a fiery cataclysm of doom.
Re: (Score:3, Informative)
Sad but not surprised. (Score:5, Informative)
I can kinda understand where they are coming from here even though I hate features getting removed. It would not surprise me if they had some pressure put on them from manufacturers. I have a friend who has basically killed two laptops overclocking them, he then takes them back and demands they are faulty, I am sure he isn't the only one doing that. Most stores don't have the technical people to be able to identify the cause on the spot so they accept the swap, especially if it is in the first few weeks of purchase.
Comment removed (Score:5, Insightful)
Re:Sad but not surprised. (Score:5, Informative)
I have a friend who has basically killed two laptops overclocking them, he then takes them back and demands they are faulty
Your friend is kind of a selfish dick.
yes he is. But that is hardly a rare condition. He justifies it by saying they wouldn't put the feature in their if it wasn't meant to be used, therefore it is their fault.
Re: (Score:1)
I see what you mean. Ouch. So people actually demand that NVidia prevent overclocking as a gambit, hoping that NVidia doesn't do that and then using that fact as a personal justification for defrauding stores? This is, as they say, why we can't have nice things.
Re:Sad but not surprised. (Score:5, Interesting)
No I don't think he really believes what he is saying himself. I am almost certain he will bitch and whine about how Nvidia are now taking something away that he paid good money for. He is a nice enough person about most things but some areas he is just a selfish prick and doesn't believe screwing large companies out of money is hurting anyone so he always finds a reason to justify it.
Re: (Score:3)
He can always stay on the version of the software that he "paid for" - as in, the one that existed when the hardware was released.
Its not as if he's entitled to further updates.
Re: (Score:2, Interesting)
Re: (Score:3)
That seems fair. Screwing companies out of money should be the customers default position. After all, the companies' default position is screwing money out of its customers.
So you think you should shoplift as much as possible from grocery stores? Hint: They'll just all increase prices to make you pay more, the only ones who wins are sellers of anti-shoplifting devices, cameras, guard companies and such. Same thing With illegitimate returns, the only thing his buddy is doing is pushing the cost of his own fuck-ups over on everybody else.
Re: (Score:2)
So you think you should shoplift as much as possible from grocery stores? Hint: They'll just all increase prices to make you pay more
No, they'll make you and me pay more. The shoplifter's price doesn't change.
Re: (Score:2)
Re: (Score:2)
To his credit in some cases he is quite right at overclocking. The general idea is to crank it till it breaks and then dial it back slightly.
We're no longer in the days of people frying eggs on heatsinkless AMD Athlons. One should expect that in the world of high-end computing cards implement features to protect themselves. Thermal throttling is the norm and if a vendor provides a feature to tweak performance I would say the onus should actually be on that vendor to ensure the system can't be broken.
Bitchin
Re: (Score:2)
Re: (Score:2)
No arguments on the dick part, but if the vendor-supplied software has basically an "overclock" tickbox built right in (some videocard driver packages have this), he may not be 100% wrong to take them back under warranty. Under US law at least, there are implied warranties for "fitness for a particular purpose", which a company cannot always disclaim once that company has implied them (this varies by state to some extent). A good example is a pickup truck that comes with a ball hitch and is shown in the TV
Re: (Score:2)
I have a friend who has basically killed two laptops overclocking them, he then takes them back and demands they are faulty
Your friend is kind of a selfish dick.
yes he is. But that is hardly a rare condition. He justifies it by saying they wouldn't put the feature in their if it wasn't meant to be used, therefore it is their fault.
He has a point; absent the manufacturer saying not to overlock or the warranty is void. Is it any different, in principle, than a manufacturer letting you run the CPU at a higher clock rater in "Turbo" mode and if it fails saying "tough luck?" I can see why Nvidia pulled 2015-02-16he capability since they probably never intended for it to be used because it can cause problems and leave them open to repair claims.
Re: (Score:2)
Re: (Score:3, Insightful)
After NVIDIA's refusal to step up to the line and assume their financial responsibility for causing so many laptops to die of thermal stress in previous generations, I feel no empathy for them. I DO have empathy for the rest of the people in the supply line that are getting dicked over by your friend, just not NVIDIA.
Re: (Score:1)
Re: (Score:1)
You are free to rest the fault with Nvidia, but at least get the fact right...
* the lead free solder was used by nearly *everyone* in the industry (not just nvidia), it is the same as the Xbox RROD which used a microsoft/ati graphics chip and was an industry issue first discovered by intel on their motherboards. As with most fabless semiconductor companies, Nvidia used one of the two main subcontractors that are responsible for about 85% of all chips "bumped" in production. Everyone used the same stuff.
*
Re: (Score:2)
It's not just nVidia. Many modern devices suffer from cracked solder joints. I found a big projection TV 2-3 years ago on the curb, the convergence was all screwed up thanks to broken solder joints between the flyback and motherboard. All it took was removing the ROHS crap and resolder using *standard* solder. Still works to this day.
So by using this kind of solder and being able to say they're environment friendly all it does is put MORE electronics in landfills.
Re: (Score:2)
After NVIDIA's refusal to step up to the line and assume their financial responsibility for causing so many laptops to die of thermal stress in previous generations, I feel no empathy for them. I DO have empathy for the rest of the people in the supply line that are getting dicked over by your friend, just not NVIDIA.
Well, they are resolving the responsibility issue by stop you from overclocking...
say hello (Score:1, Funny)
-----------------
"fcuk Nvidia" -- Linus Torvalds
Just wait for Carpel Tunnel to set in. (Score:2)
Just wait for Carpel Tunnel to set in. It is like someone turned off the turbo on your computer/hand.
The Point? (Score:1)
Really, you need to OC the GPU on a freaking cheesy laptop?
What, are your Flappy Birds flapping too slow?
Overclock on a laptop? BBQ! (Score:5, Insightful)
Re: (Score:2)
The 980M in my Clevo P650SG overclocks by 125MHz with ease - and it won't even hit 70C while playing games in that overclocked state either. When you're playing at 3K (there's also a 4K screen available), that extra 125MHz makes a noticeable difference.
Removal of overclocking from the drivers is irritating at best.
Re: (Score:3)
What about the people living where water outside is not in a solid state at the moment. BBQ!
Re: (Score:1)
The cheap and nasty Acer will throttle when it gets to 70C, overclocked or not.
The water outside here isn't frozen. I'm in the southeast of the UK, where we've had a generally mild, largely snowless winter - it's 8C as I write this, for example. Not that that matters, as most of us have central heating and the temperature indoors won't be anywhere near as cold as it is outside!
The 2nd-gen Maxwell chips are known to run cool and overclock well, be they laptop or desktop form (in fact, it's the same silicon -
Re: (Score:2)
A cynic would say they're removing overclocking as it'll impact on their plans to release slightly faster versions of the same chip later this year...
That would almost make sense if we were talking about desktops, where the user can replace the GPU. That's typically not an option in a laptop; even if the GPU is designed to be replaceable, it's oft not designed to be upgradeable, or the manufacturer doesn't make parts available. And I say almost because anyone who cares about having the fastest laptop GPU on the block enough that they overclock their laptop GPU will run out and buy one of the new models that has a GPU that comes from the factory running a
Re: (Score:1)
FWIW, the sort of laptops for which this is an issue (ie high-end gaming laptops) typically have graphics cards on MXM modules, they're designed to be upgradeable to the latest and greatest.
The *price* of those modules, however, means it's not something most owners will ever do...
Re: (Score:2)
Re: (Score:2)
It would be fine if they were simply removing OC support on new cards. The problem is that they are removing a feature from a product that has already been sold, and even used as an advertising point by some manufacturers. This is no different to Sony removing OtherOS from the PS3.
Understandable (Score:4, Insightful)
NV-using laptop manufacturers forced their hand (Score:5, Insightful)
This is all about warranty repairs.
Less than bright people overclock their laptops to unsafe levels, laptop dies after 6-12 months and ends up on the laptop manufacturer's repair table. No way to conclusively prove it was overclocked, so they end up picking the tab for the hardware abuse.
Laptop drivers have allowed overclocking for a good while, so it must be that some recent generation NV chip had unusually tight margins and there is a noticeable spike in warranty claims, or just some big laptop manufacturer not wanting to deal with the headaches of overclocking-related support/warranty incidents is suddenly pushing NV to solve he issue on the driver level or lose business.
High end GPUs have always had fairly tight thermal margins. Even more so on laptops. The age old problem of packing really high performance silicon into laptop form factor with tiny heatsinks and small fans. Sure, they could just downclock the chip by 20% and have a nice, cool laptop that... would lose to the competitor GPU and really mess up the sales of the chip. So they push it as far as they possibly can... and the tight margins on laptops just can't do any meaningful overclocking without completely replacing the cooling - which is not really doable in a laptop.
At least on NV side you generally can always install the "generic" laptop driver and get the latest driver bits. On AMD side there are many laptop manufacturers that outright block the generic AMD graphics drivers for ~reasons~ and you end up with a piece of hardware that has effectively an unsupported GPU - laptop manufacturer cannot be assed to update the GPU driver and generic drivers do not install (unless modded).
Surprised they don't want them to fail! (Score:4, Insightful)
Coming from the people who brought you self-destructing laptop GPU chipsets and fan profiles that prefer quiet to reliable.
Does no-one remember the 8800M GT?
Damn fools used low melting point solder on them and they failed worse than Windows ME.
They never have and never will specify decent cooling.
Many a fine laptop has been turned into a brick by this short-sightedness.
But never mind, you can just go buy another!
Re: (Score:3)
What a load of drivel modded insightful by the uneducated. I like a good conspiracy theory as much as the next guy but this just seems like someone read the first explanation on toms hardware and declares themselves an expert.
The problem was not the melting point of the solder. It was fatigue cracking. The switch to a different type of solder has tradeoffs, everything always does. The same solder was used by both Nvidia and AMD at the time. In Nvidia's design of a series of GPUs of the same age someone got
orly? (Score:2)
Can I still underclock my GPU? (Score:5, Interesting)
I had a dell XPS15 model with a manufacturing defect that made the computer crash when running games, the only way to solve it was to underclock the GPU. No it was not overheating, it was a manufacturing problem, there are several accounts of this problem on the net.
Risky move for NVidia, let's see if it pays off (Score:2)
I am sure this is based on some analysis of failure data. Regardless, this is a bad move when people are already cooling off on discrete graphics, especially on laptops. Intel integrated graphics will now run many games adequately on small screens and there are obvious cost/form factor/battery life advantages. If you don't cater to hardcore gamer/technology enthusiast market that is most interested in overclocking, just who is going to buy your chips and cards?
Re: (Score:2)
If you don't cater to hardcore gamer/technology enthusiast market that is most interested in overclocking, just who is going to buy your chips and cards?
I would guess that market is small enough not to worry about any impact on sales. If including discrete graphic chips is important for mainstream sales manufacturers will use them in some higher end systems; if not then they are on the way out wether or not they can be overlocked.
Sounds like a good use for FPROM? (Score:2)
From a technical standpoint, it seems like the ideal solution is to have some programmable ROM that users can blow to indicate that they have accepted any harm that comes from clocking it beyond what the design (heat/voltage/lifetime) allows. That ROM would have to be queryable via a tamper-proof BIOS or EFI hook so that stores could verify that it is intact before accepting returns.
Ultimately, user freedom to do what they want with their own hardware has to come with user responsibility over the consequenc
Re: (Score:3)
And?
They supply chips to manufacturers and overclocking is specifically RUNNING SOMETHING PAST ITS DESIGN SPECIFICATIONS. As such, overclocking causing product returns is exactly the kind of things that suppliers will push back to nVidia on to cost them money.
How would you like it if you designed sold a laptop for which a component manufacturer allowed you to ramp it up to 120 degrees when your case was only designed, built and tested to withstand 90 degrees? Every return you get, you'd either push back t
Re: (Score:2)
Lemme guess? eeePC? If so, it didn't run at 400. It just said it was running at that speed...