Bitcoin Forum

I got a bunch of xfx 5870 with broken fans. After failing to find proper 12v x .5A replacements, I decided to use 12v x 0.8A CPU fan in a piggy back mounting configuration.

Problem is, when connecting their 4 pin connectors to 3 pin sys fan sockets on mobo, they always run at 100% (very noisy too!), even though the cooling is excellent, I have to wonder how long they will last.

First, don't worry about the amps.  You want a 12V fan, pick the one with the highest CFM (and/or the lowest db, depending on your needs).  plus or minus a few Watts won't make a damned bit of difference.

That said... The speed of a cooling fan may slightly decrease its useful life, but basically speed doesn't matter except for noise level.
 
They also don't tend to die while on, only when spinning up.  So for a 24/7/365.24 machine, you really don't need to worry about it.  Seriously, looking at 5+ years even at full speed before it dies (somewhat less if you have pets, and halve that if you smoke indoors).  As it gets near the end of its useful life, pay more attention when you power it up to make sure it actually starts (they usually start making a buzzing/grinding noise for a few seconds on startup, for a few months before completely failing).
 
As for the 3 vs 4pin connector, you've chosen the "wrong" 3 pins.  You may need to actually split out the wires to get it right, buy you want R/B for +12V/Gnd power, and yellow for the fan signal.  That said, fans cost nothing while good GPUs don't come cheap.

LOLZ.

[/quote]
I do hope so. but considering 5870 has only been out 2 1/2 years and I have a handful broken fans, I am a little skeptical.

It's hard to believe fan speed has little to do with its lifespan, unless the bearings are magnetically suspended and frictionless.

The fan in an reference ATI video card might last five years of normal use in a CLEAN COOL environment, mining 24/7 is not normal use.  Even normal clocks they simply will not last five years running constantly mining..   These fans seem to die based on how long they run over 50% speed, so one card on its own not overclocked will run a while, maybe years.  Add in either overclocking or high constant heat from neighboring cards and you are looking at higher fan speeds and earlier death.  

I do not believe you can run most reference cards one year at full speed, let alone five years without failure.  

A proper ball bearing fan should last a really long time.  It's the cheapo sleeve bearing fans placed on GPU's that fail when pushed hard.  They were designed for someone playing a video game a couple hours a day, and even then the PWM shouldn't go above something like 60% based on the load a game places on a modern GPU and ambient room temps in an average house.  I should point out that *most* fans you buy, for computers and all other applications (think about fans that run on 110V AC power) are running at 100% all the time.  I don't buy a big fan and go "Hmm, what should I set the speed to in order to make the fan last longer".  What should be focused on is the quality of the bearing instead of what percentage the PWM is at.

I have HD5970s that have been mining 24/7 with the fan at 100% for 1+ year. My oldest one has been mining for 1.25 years. But then again I filter my air. There is barely any dust visible on the fan blades. They wouldn't last that long in a dusty environment.

Also, non-reference cards sometimes use fans of quality inferior to the ones found on reference AMD cards such as the HD5970.

Excellent point - Consider my previous post as not applying to cheap sleeve bearings.  When buying cooling fans, don't skimp on the extra $0.50.

Interesting, probably the longest lasting 5970 fans in history, if you have actually been running them at 100% the whole time. What is the card orientation?

Or just variance.  I got some cards going on >1 year.  Although out of 24 5970s I have had to replace 4 fans so far and not all of them have been running a year. 

I'm curious how do you filter the air?

Then I should consider myself lucky, they are "double ball bearings", a lot of balls. 

I don't think mining 24/7 on HD5970 necessarily means 100%, they rarely reach full speed.

If you want to know how loud 100% is, plug the cards in PCI-e without connecting those 4/6 pin power cables, they are VERY noisy.

I love that marketing gimmick on small fans, lol.  They act like it's an industrial HVAC fan with a foot long shaft that needs to be supported on both ends or something.  Might as well say something like "DOUBLE BALL BEARING TECHNOLOGY! -- Our 120mm fans have a huge shaft that requires a lot of balls"

yeah ~~ but it always feels good knowing you have a lot of balls under your belt.  ;D

They seem to have many varieties for case fans. Anyone have experience with some of the more exotic constructions? Rifle, liquid, & maglev, for example?

The cards are vertical (motherboards are horizontal).


4 out of 24 is a 17% failure rate for an average period of less than a year... For comparison my AFR (annualized failure rate) for fan ball bearings is 7%... far from the "OMG you will kill your cards in 6 months if you run the fan at 100%" FUD you find on the forums.


With air filters on the air ducts going to the server room.


I actually do force the fans to run at 100% with aticonfig.

Wow, looks like you did all the right things -- hats off to you Sir.

Now I almost feel ashamed of the atrocity I've committed to my workers ;)

17% failure rate,  NOT running at 100% fanspeed.  I would say that does uphold the 'fud' about killing fans within a year running 100% fanspeed. 

But ok... if you want to run your fans at 100%, go ahead.   Miners offline=lower difficulty.  Fans dead=chance of purchase of replacement from me.   ;D

I had to laugh at that too.  The things you read on this forum, just hilarious.  Everyone is a professional.

My lower 7% AFR at a higher 100% fan speed proves you wrong.
My point is that dust, not fan speed, is a much bigger factor affecting ball bearing life span.

Dust, dust, dust. It's all about dust.

7% isn't 0%.    Now if your were on year 5 with 0% failure rate then maybe it proved something.

Considering that it is 17% for the reference fans, the ones most of us are stuck with, seems like the figure I am going to use.  And on those fans it is speed as well as dust that are the factors that kill them. 

Mine are reference fans as well. All HD5970s are reference designs.


7% still proves my point as it is significantly inferior to the 17% of someone who doesn't even run at 100% fan speed. Besides, why would you want to mine with a 5-year old card that consumes 4 times the energy to produce the same amount of work? I expect to upgrade my 5970s after about 2-2.5 years. They are already more than halfway through their service life.

I think you guys also forget that skimping on cooling and running a card 10C hotter will wear out the electronic components faster.

I know, right?  Lookit this moron, who answered the initial poster by telling him the right pinout to properly convert between 2/3/4 wire fans off the top of his head, yet seems blissfully unaware of the high fan failure rate experienced by those unwise enough to run a stock bottom-of-the-barrel cooling fan.  Whadda maroon!

I do have to agree with you about the truly... "incredible"... level of expertise on this forum, however.

to avoid malfunctions, gpu fans requires a degree of savvy and caginess that for many may take years to master, if ever

one should seek assistance from a professional before prematurely attempting operations

??? ??? ???
https://i.imgur.com/22yHR.png

Being sarcastic, or trying to be?  :-\