Slashdot Mirror
AMD: No Grease For You!
bahamat writes "In a surprising turn of events, this article over at Xtreme Tek explains that the official stance from AMD is that you will void your warranty if you use any thermal grease or if you're not using the heatsink provided with your CPU. Sucks to be you if you buy a defective AMD CPU and put a Zalman on it for the first boot." AMD, the article says, doesn't want you to use anything "other than Shin Estu G 749."
So what does that say? Either too many people are either applying the grease too thickly, or people are using heatsinks and OCin' their processors too much.
Basicly what it sounds like is, people are cooking their CPU's from either/or...and AMD is tired of floating the loss because of idiots.
Om, nomnomnom...
Actually, they could very easily know. All they have to do is put some kind of alcohol-soluble, UV ink mark on the chip. It doesn't even have to be visible in UV... that's just if they want to be sneaky about it.
-- Minds are like parachutes... they work best when open.
This isn't necessarily a good thing, especially when you fuck over businesses in order to do it. If you don't agree with the business' stance, and want to use a different heatsink while keeping your warranty, but Intel. If you think it's worth voiding the warranty, by all means, go for it. Don't force companies to cater to idiots who don't know how to install hardware properly.
No comment.
AMD does not have the legal right to prevent you from using 3rd party heatsinks as long as they're designed for the AMD CPUs. This is the same as saying using 3rd party ink in the printer will void the warranty. In both cases, the company is still legally obligated to honor the warranty, but fighting them in court for it is another matter.
Jason
ProfQuotes
The Shuttle SN41G2 requires a user to not use the AMD heatsink, as the box comes with a custom heatpipe to get the heat from the CPU away from the chip, and the insides of the system. Using the AMD heatsink could lead to the system building up too much heat internally, and causing the box to shut down or crash. I'm sure AMD processors are also used in similar custom machines.
Why is it the consumers fault AMD never integrated thermal protection or a heat spreader into their processors to protect them? Personally, I'm glad I have a P4 in my gaming box that won't fry its self if a fan dies.
uh, because with their thermal crap and the defauult fan, my proc ran over 130F. With artic silver and an aeroflow fan from vantec, my proc runs at 105F (115F under heavy load). Seems to me they are just trying to save a few bucks, afterall, the german economy has taken a hit lately. That is the problem here
YOU SUCK BALLS!
Sorry to interrupt you tirade, but has it occurred to you that overclocking probably also voids the warranty regardless of what heatsink you're using? Just a thought, you know...
or with a dangerden kit it wont break 40C, like mine!
I don't see a problem with this. AMD will only guarantee its equipment with parts they provide. Use your own parts, and you're responsible for damage. Sounds perfectly reasonable to me.
"Jesus saves, but everyone else in a 10 foot radius takes full damage from the fireball."
I've installed my last three heatsinks, including my current Zalman CNPS6000AlCu, with pink TIM wax pads. TIM pads are cheap, neat, and don't require you to get crap all over your hands during application. I'll never go back to the goop game.
Since it doesn't make a difference what the hell you use to stick your heatsink onto your CPU--hell, toothpaste works just as well as AS-3--I'll stick with the easy stuff.
I have two AMD Athlon MP 2000+'s in on a Tyan Tiger MPX motherboard, and a gig of ram, in a full-tower case with four intake fans -- one on the bottom front, one on the side middle over the cards, and two in the middle back under the power supply. The exhaust fan is the PS, of course.
When running Windows 2000 on this machine, the operating temp as reported by the BIOS runs between 50c and 60c.
When I run Gentoo Linux, set up from a stage1 install and compiled specifically for the Athlon MP, the machine crashes as the temperature rises to 75c.
I'm using the retail processors that came with the fan. It's plain that they're installed correctly, but the thermal pad on the bottom (even with the adhesive backing removed properly) isn't capable of dissapating the heat.
Does this mean I'm prohibited by warranty terms from running optimized code? AMD really needs to answer this question. If they want to they could easily come up with a recommended list of approved grease, or contract with someone to sell "official grease" for situations like mine.
Remove the caps and hold to a mirror.
Exactly where did AMD say it would void the warrantee? All AMD said, according to the article, is that they recommend a particular type of thermal grease. They didn't even MENTION the word "warantee"!
The author is jumping to conclusions prematurely.
Jason
No one has seen what you have seen, and until that happens, we're all going to think that you're nuts. - Jack O'Neil
1) using grease means you are DIY with 3rd party stuff, which means it's easier for you (or the 3rd party suff you have) to screw up and
2) using grease *improves* thermal contact, making it easier to overclock, which of course voids your warranty anyway.
Now compare the XP's "no grease" tag like to this from the AMD Athlon(TM) 64 Processor Thermal Design Guide (from page 22, secion 2.6.6):
For those that don't know, the gum-paste stuff that comes on all XP heatsinks is "phase-change material." Seems the 64 is the *exact* opposite of the XP.
Do they support Palladum and other restrict-what-I-can-do-to-my-PC technologies?
Zodiac Survey
This is a very inexpensive, easy to implement overclocking deterrent.
/., I can understand why they'd do this. There's no way for them to tell if you weren't using a HSF on their approved list.
Think about it. If it's a poor heat sink and fan that AMD requires you to use, then overclocking is out of the question because there's no way the HSF could regulate the temperature.
See, overclocking voids the warantee, but there's no way for AMD to tell if you were overclocking. This way they do know (or it's easier for them to tell) - assuming you used a different HSF while overclocking. I wonder what they'll do when hot cpus start failing in poorly ventilated cases, or in hot climates.
Personally I don't like it, because a crappy HSF... just sucks. I don't overclock, but I still want a half decent HSF. You could probably get away with using a different fan though. Watch, soon you'll see "heat sink extensions" that you lock on to AMD's required sink, and it extends the surface area of that heatsink.
Though given some of the crazy cooling solutions that have been posted on
-kidlinux.
Damn straight! I wish I could mod you up ! I was working at this electronics company, and there were these linear voltage regulators running REAL hot (10W in a T0-220, no joke), and the tech smothered them in grease, I mean thick gops everywhere, PLUS those rubber spacers. So the case gets barely warm and he thinks he's the thermal champ.
Of course, when I saw that shit I blew a fuse, (it was my design after all), but I couldn't make the guy grasp that the case has to be HOT for the heat to sink away!
Why exactly would I be voiding my warrenty by switching to something better?
There is someone who apparently works at AMD over at the Bit-Tech forums who has been giving out unofficial information that seems to have quite a bit of merit to it. So make sure to check out what he is saying as he knows quite a bit about what he is talking about: Bit-Tech Forums (Remember, what he says isn't official, though)
I am suprised that no one has mentioned "The Magnuson-Moss Warranty Act" which could be taken to forbid this kind of warrenty restriction.
It is the same law which forbids printer manufacturers from voiding a warrenty when you use a third party printer cartridge.
I suspect that AMD could forbid all thermal greases but limiting it to a specific brand probably would be illegal.
" Actually, they could very easily know. All they have to do is put some kind of alcohol-soluble, UV ink mark on the chip. It doesn't even have to be visible in UV... that's just if they want to be sneaky about it."
And when they say you cleaned off your artic silver you tell them you used the alcohol to clean of the approved grease so the chip was nice and clean for them.
130 F is only about 54 C. AMD's processors are rated for a max of 90 C (the XPs are anyway, it varies slightly from processor to processor, but that is ballpark). 130 F is not at all dangerous (assuming you were actually measuring from a thermal resistor, and not from a thermistor under the core). The fact that you got the temperature down to 105 F (about 40 C), is simply a don't care from a part lifetime point of view. Most likely they are designed to last 10 years at 110 C (maybe 125 if they are more pessimistic).
What you want to do is go to a law library and look up Vandermark v. Ford Motor Co. 61 Cal.2d 256 (1963).
Vandermark bought a new Ford in Los Angeles. Six weeks later, with 1500 miles on the odometer, the brakes failed, causing the car to wreck, seriously injuring the driver and passenger.
The Ford Dealership acknowledged that the crash was caused by defective brakes, but pointed to the warranty that read, "Dealer's obligation under this warranty is limited to replacement ... of such parts ... acknowledged by Dealer to be defective." In other words, neither the dealer nor Ford would assume responsibility for the damage to the car or the injuries to its occupants caused by the defective brakes.
One could argue that if Vandermark wanted a car with a warranty that would cover defective manufacture more comprehensibvely, he should have bought a car from a dealer that offered a better warranty, but no dealers offered warranties with significantly greater coverage. In 1964, the court ruled that the Ford dealership was strictly liable for product defects irrespective of what the warranty might say because
. This followed the train of thought set in motion by Benjamin Cardozo, who wrote in 1916 in MacPherson v. Buick Motor Co. 138 NYS 224 (1912):You may argue that it is a long way from automobiles whose manufacturing defects put consumers and bystanders in danger of life and limb to a defective cooling fan on a CPU, and you would be right. But if you complain in general that the state has no business interfering with product warranties, a century of case law disagrees with you.
In closing, I will point to one of the most egregious cases in this regard. In 1937, the Massengill Company put on the market an antibiotic elixer for children composed of the drug sulfanilamide dissolved in diethylene glycol and flavoured with raspberry extract. Massengill never tested the product for safety. Diethylene glycol being a very nasty poison, 107, mostly children, died shortly thereafter from liver failure caused by this medicine. Massengill could not be sued under the laws at the time because, as the President of the company said,
The nation's response to this was to pass the 1938 Federal Food, Drug, and Cosmetic Act, which established the FDA and granted it sweeping powers to regulate the market to ensure that all food, drugs, and cosmetics were safe. Many conservative jurists, such as Richard Posner (one of Ronald Reagan's first appointments to the federal bench) promoted this expansion of tort law, noting that there is an imperfect market for information and that when information asymmetries are present, a free market does not optimally allocate resources (this observation won a Nobel prize in economics for Ackerlof, Stiglitz, and Spence). The thinking of the economics-and-law crowd was that expansion of strict liability would produce a corrective force for disclosure of information that would enhance the efficiency of markets.If the ACME supercharger at a minimum meets the same engineering specs as the TRD approved unit. Then Toyota may very well be legally obliged to honour their warranty regardless of their written warranty limitation.
Consumer protection laws allow the stipulation of specifications, not brands, models or arbitary approval processes.
As always it depends on your local laws. The manufacturer must put up, or leave.
I bought a couple before, they worked fine. After of course some heatsink tweaking to stop it crashing all the time. Then I found this statement somewhere:
"Be very careful if you want to install AMD CPU's yourself, because the AMD warranty policy became very severe. Guidelines from AMD: '' NO warranty on mechanical damage of the CPU like bent or broken pins, cracked dies or packages. NO warranty on manipulated CPU's by e.g. overclocking. No warranty on overheating e.g. caused by use of non-recommended (www1.amd.com/athlon/config) or improperly mounted coolers (www.amd.com/products/cpg/athlon/pdf/23986.pdf). or by overclocking or use of core voltage different from the datasheet. You can recognize overheating by e.g. brown spots on bottom side of the CPU, different colors on the chip surface, destroyed support pads, head conducting pastes spread all over the package. Especially no warranty for findings like: open/short at Vcc- or I/O pins, Micro Cracks (invisible cracks), when part is fully functional. ''"
Why is their warranty very severe? And why are they backing off thermal grease and non approved heatsinks? Because they have a severe overheating problem, and their design is certainly not rugged enough to bear any type of slip or mistake. With some investigation I realise that most of my friends with AMDs have overheating problems, one running 10 fans in the case to get it 55 degrees.
You get what you pay for, and AMD is cheap. Their stock heatsinks are crap, and some of the design decisions they've taken are flawed. Ho hum.
Its sad, but next one for me is a intel. At least it might still be running on a sunny day, rather than setting fire to the house...
Toothpaste was actually better at transferring the heat than arctic silver. (Not that toothpaste would be a good solution in the long run :))