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."
State Law mandates a 1 year parts and manufacturing warranty irregardless of OEM status or not. Say what you will about Calif. but we do look after the consumer, except for the power consumer that is...
"Sucks to be you if you buy a defective AMD CPU and put a Zalman on it for the first boot"
AMD Athlon XP(TM) Processor Thermal Solutions
When selecting a thermal solution for your system, please refer to the following listing of heatsinks evaluated and recommended by AMD for use with AMD Athlon XP(TM) processors. Note, the following heatsinks are recommended for maintaining the specified Maximum die temperature requirement. In addition, this selection is not intended to be a comprehensive listing of all heatsinks that support AMD processors.
From the article:
"using a heatsink other than the stock heatsink on a retail chip also voids the warranty."
So if you pay the $10-15 extra for the retail instead of OEM, they expect you to use the hs/fan that comes with it.
Whenever the offence inspires less horror than the punishment, the rigour of penal law is obliged to give way...
well, i can attest that people do use too much grease, i know i did when i first put on my own fan. I corrected the mistake, and the proc has been fine since. my question is, if you can't use grease, what do you use. that foamy pad stuff? it doesnt work nearly as well. Either way, i'll stick with my after-market fan and arctic silver, warranty be damned.
YOU SUCK BALLS!
Didn't any of you notice that they only specifically said that Arctic Silver will void your warranty? The reason for this is that Arctic Silver is a Thermal EPOXY... While the text of the article also points out that they don't like the "conductive properties" of (obviously) the minute ammount of silver in it, they also on their "green-list" had a thermal greese that does have minute ammounts of silver. I do not think they are trying to dictate what you use, beyond trying to keep idiots from costing them needlesly large ammounts of money. Face it, just about everyone in the world thinks they know enough to set up a computer, but when they do it, they will completely forget the HS/FAN or use too much thermal greese to the point where it's minute conductive properties cause shorts... It is the idiots they are after, not us... give them a break!
Erutangis ym si siht.
The ultimate cooler review site: http://www.dansdata.com/coolercomp.htm
:-)
Just a good one to bookmark
No, I did not read the f***ing article!
Let me explain the problem with a simple example. And this has nothing to do with those who want to overclock their CPU or otherwise soup up their system.
The fan on the "stock heatsink" they talk about is less than quiet, but more importantly is poor enough that in many cases it will not even last the life of the warranty on the CPU. And the phase change material is "one use", you can't remove the heat sink and reapply it again with the same strip of heat sink material and have it function properly. When my heat sink fan died some months after I started using it (as detected by the BIOS seeing it slowing down considerably and a reelated increase in CPU temperature), I went to the local CompUSA and got a replacement copper fan. The "stock heatsink" just isn't always available. The replacement heatsink does a far better job, but I had to use a non-conductive thermal grease to install it. By this proclimation, I would have voided my warranty in two ways, using thermal grease and a better but non-standard heat sink.
I much prefer AMD over Intel, but if AMD is going to do this they need to consider some real world situations. As far as I know they don't warranty the heatsink fan, and even if they were to start it would not be reasonable for them to expect a user to not use their computer for the time it takes to ship back a bad fan and get a replacement. If they were supplying a fan and heatsink that would never need replacement, they might be on a more moral high ground, but having a stock fan that dies easily and then claiming you void the warranty if you correctly replace the heat sink isn't user friendly. And, of course, there are some people wo have a problem with the noise the stock fan makes, and while the argument is not as strong as the one I just made, I think they should be able to replace the stock haetsink and fan with a quieter one if they want, as long as they follow good technical procedures. I certainly don't have a problem with AMD stating they will not be responsiable for problems caused by conductive thermal grease, but this policy does seem to go too far, particularly given the fan they supply on that heatsink.
I'm an American. I love this country and the freedoms that we used to have.
Rubbing alcohol! They'll never know.
No, try again. Ever tried getting Arctic Silver (or el-cheapo white thermal paste) off a CPU with isopropanol and a Q-Tip? You won't get all of it off.
So what is the secret sauce you need to soak that Q-Tip in, you ask? Acetone. Gone without a trace every time!
I recently bought an Intel retail boxed P4 and they strongly urged you not to use thermal grease. They said instead to use the thermal tape that came on their heatsink, and I did. Their reasoning is that the grease dries and it makes it impossible to remove the heatsink from the chip and you'll damage the chip.
scott
If the CPU was so covered in grease, it's no wonder it ran hot. Thermal grease is meant to be used very sparingly. You want a very thin layer, just enough to fill the microscopic crevices on the CPU and heatsink mating surfaces (if the surface of the heatsink has larger-than-microscopic grooves, it's time to get out the sand paper and lap the base to a shiny finish). As good of a thermal conductor as thermal grease may be, it's terrible in comparison to a direct metal-on-metal connection. If you see a lot of grease squeeze out when you apply the heatsink, you've got too much. What you want to do is apply a thin bead of grease to one end of the CPU die, and spread it thinly across using a credit card or other device (driver's license, heavy piece of paper, etc).
If your thermal grease is applied correctly, it will come off of the CPU with little problem. A cotton swab and a dab of rubbing alcohol should be more than enough. If you have to scrub, you had too much grease to begin with.
I should know I've installed several dozen and got pissed at using various custom substances so a qtip dipped in wd40 along with a soft tissue dipped in likewise and one piece of toilet paper (to wipe off that nasty ass gasoline smell it leaves behind and the greasy looking stains you'd otherwise have left on it).
:)
And voila, looked like new, took a new coat of thermal paste/grease and... ran like new, all my servers still running after 2.0 years
That's my 2 cents.
JOKE: And to boot I bet it will never squeak either. Plus WD40 cuts through roughly any household thermal grease, including the vaunted Arctic Silver II stuff. The silicone paste comes off with water if anyone's worried. A moist towelette type thing will work just fine including the Staples/Radioshack anti static wipes. Hope that helps.
-Daedalus
PS - Arctic Silver 2 is good stuff, but not really much more effective than say... radioshack 1.99 tubes of thermal goo... the only rule is that you should use SOME form of evening compound to fill in gaps and uneven surface finish between cpu and heatsink... any paste will usually do, even thermal pads are better than nothing, unless that ONE degree difference from Arctic Silver II REALLY means much to you... in general all the silver compounds do is react to temperature changes more quickly... but unless you're running a superspiking cpu, you won't have an issue, and I have never had a cpu that spiked so fast and so hard that the thermal paste made a difference. Oh and to add, I've never had a CPU fail. It is all about cooling and how often you turn them on and off. Much like diesel motors...
" What luck for rulers that men do not think" - Adolf Hitler
Actually, your example is true. Japanese automakers will not warranty engines that have had silicate-based antifreeze put in them. German cars typically don't allow phosphate-based coolant.
Why? Because silicates cause increased wear on moving parts, like water pumps. Water pumps are very often driven by the timing belt. When the water pump seizes, it usually shreds the belt in a matter of seconds, and if the engine is an interference type(ie, the path of pistons and valves overlap, but never hit because of the timing), then you'll bend/snap valves, or worse.
Phosphates don't react well to water with mineral content(US coolant makers claim it's only a european-water problem, that the US doesn't have "high mineral content" in its water, which is bullshit), aren't friendly with aluminum engine components/radiators, and like silicates, they work by coating all the metal with the stuff(on the theory that, if a metal that can rust is covered by phosphates or silicates, it won't rust.)
I use the proper coolant that was recommended by my car's manufacturer- it's german-made, and doesn't contain either phosphates or silicates. I've actually seen better operation(less noise from the water pump, for one) since I switched.
The reality is that if you substitute coolants, and your engine overheats because of it- you're shit out of luck. That said, the manufacturer has to prove(to a certain extent) that your coolant switch caused the problem(which could be as simple as "see this pump? It seized because you used crap coolant.") Auto manufacturers can't just declare the whole vehicle's warranty invalid because, say, you install a non-OEM air filter.
Maybe it's just me, but computer manufacturers have clung to the "open the case, void the warranty" bullshit. Some invalidate the warranty because you installed, say, a network or faxmodem card. That's bull- just like auto companies, they should be forced to prove the non-OEM component caused the failure.
Please help metamoderate.
It DOES work- and well. But I've found that "Goof Off" and "Goo Gone" do an equally adequate job and don't have anywhere near the flash point acetone does...
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
The cheap stuff (including the stuff AMD's reccomending) contains oils that tend to dry out in a year or two, leaving this gummy mess that won't come off with isopropyl. (This is the main reason for my using Arctic Silver- it doesn't dry out like Shin Etsu, etc.) For that, you've got to use something like acetone, WD-40, Goof-Off, or Goo-Gone- even if you've applied it correctly.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
btw, do NOT use acetone to clean your credit card after using it to spread the thermal paste... as I said before, acetone cuts through almost anything, including the surface of credit cards. At least mine still has a magnetic stripe.
Acetone = good call :)
One easy way to obtain acetone is from nail polish removers.
Great idea. If you want to dissolve half of the CPU package, that is. DON'T USE ACETONE ON ANYTHING THAT REMOTELY RESEMBLES PLASTIC, including CPUs.
It's a pretty low risk material. Alot less toxic than motor oils, paint or glue. Don't rub yours eyes, lick your fingers, or eat the stuff and wash your hands promptly and you'll be fine.
Here is a hazardous material datasheet for a typical thermal compound.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
"You do something that could _potentially_ damage the processor"
Due to consumer protection laws, they may not void a warranty if the potential problem did not actually contribute to the actual failure.
And a heatsink of better thermal capacity using an equivalent spec non-conductive thermal compound could in no way "potentially damage" the CPU. Seems to me they are trying to weasle out of returned CPUs that have been over-clocked and died early. And why I agree that the CPU should never have been returned (and the warranty voided) if it was run out of spec, this is the wrong way to go about it.
I think they should just focus on whether the CPU has been run outside of spec. There are technical ways to one time record (true or false) over-voltage/clock/heat conditions.
They are locking the barn after the horse has bolted. I can only hope that they have learned a lesson here.
Athlon 64 has a integral heat spreader (like the P4 and opteron) instead of an exposed die. This means the core is well shielded, and has nothing exposed to short out. Athlon has the die exposed, and bridges on the top of the chip exposed, which conductive greases sometimes short out. So, the mounting configuration is pretty different, and it doesn't surprise me in the least that the recommended way to mount the heatsink is different.
The Matrix is going down for reboot now! Stopping reality: OK. The system is halted.
Yeah, you learn something new everyday. I've been in electronics for ages and never heard of it before too.
After doing a little research, it refers to the migration or "pump out" of the grease due to the pressure of the heatsink on the chip and the thermal cycles of turning your cpu off/on. Because the cpu and heatsink are usually different materials they will expand/contract at different rates, and thus will "push out" the grease.
Then this might create a space between the cpu and heatsink with the "pumped out" grease acting as a standoff, essentially leaving your cpu without meaningful contact with a heatsink. With AMD cpus that normally run very hot, this is deadly and your cpu will melt.
Thermal Insulating Materials, or TIMs have a much higher viscosity than grease, so that won't happen with them. Though arguably they don't do as well as grease (in the short term) in the thermal transfer of heat.
On the surface this seems like a conspiracy by those who make TIMs, but one user observed "pump out" for himself - using Artic Silver too.
If the AMD cpus had a heat spreader on them like many intel cpus do, "pump out" may not be so bad. Though it probably would happen to some degree, it probably would not be fatal.
Given the vast quantity of cpus that AMD (and intel for that matter) crank out, they probably see a steady stream of cpus burned out with grease on them been returned to them. So naturally they want to cut there losses and want people to stick with TIMs to prevent burn out.
It also could be argued that the socket A design is inherently flawed. Thermal grease as commonly known was not designed to operate under the high pressures that heatsink clamps put on cpus these days.
I have always dissed TIMs, but now I'll give it a second thought.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
"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."
What I don't fully understand is the gripe about those who buy in the retail box who have fans that are inadquate and need replacement? Doesn't it say "heatsink" not fan? Now I don't know about anyone else in here, but can't you just replace the fan with a series of 4 screws?
While yes, they make BETTER syncs, users who's fan has failed on the stock heat sync, who have no complaints about the present heat sync, provided their replacement fan is within the same size, speed, and airflow specifications don't have anything to worry about (in theory).
Personaly, I find it "much" easier removing 4 sheetmetal screws then dealing with the cpu clip anyday!
There is no sanctuary. There is no sanctuary. SHUT UP! There is no shut up. There is no shut up.
Au contraire! It took my 20 minutes of quality Q-Tip & RS Solvent Cleaner work last time I did one...
That was classic intercourse!
Just be careful picking it out, because some of them *don't* use acetone now.
Toyota may be legally obligated to prove that the aftermarket -charger broke the engine, but that is their only legal obligation. For example... If you install that supercharger, and a warranty covered part of the air conditioning system goes, they will have to fix it, ditto for anything that cannot be proved was affected by an aftermarket/out of spec part.
Personal example:
Lost my fuel cap. Replaced with an aftermarket... the aftermarket didn't seal the system to spec. My warranty is still good, however, I had to pay out of pocket for the computer to be reset.
Lost the cap again... yes I lose rather a few.
This time I got one from the dealer... same defect, but this time, full warranty coverage...
Food not Bombs is a nice platitude but it breaks down when you notice that the Bombees are usually well fed
Solvents are generally categorized into polar and non-polar solvents (water being a good example of the former, and benzene a good example of the latter).
It all has to do with the kind of substances you are trying to mix, or put into solution (polar, ionic, covalently-bonded, etc), and your solvent of choice... oil and water do not mix because water is unable to adequately penetrate and dissolve the long, largely non-polar hydrocarbon chains that make up most oils (the oil is more attracted to itself than it is to the water). you are correct that oil and water don't mix, and polarity is the reason. The best you could hope for would be some kind of suspension.
Emulsifiers can create the appearance of solubility... but that's another lecture.
Even if a man chops off your hand with a sword, you still have two nice, sharp bones to stick in his eyes.
There have been cases where software bugs in medical equipment killed people. In those cases, there would be strong precedent for product liability lawsuits.
Suppliers to the military are harder to sue, which is probably good news to the folks whose bugs killed soldiers when their mortar targeting software made incorrect assumptions about target altitude or when a Patriot missile targeting system's clock overflowed after 8 hours.
For further reading on software liability issues, see this Business Week article, which was discussed on /. and badsoftware.com, which surveys software liability issues from a consumer's perspective.