Would you RMA a video card that you killed thru overclocking/modding?

[3chordcharlie]

Originally posted by: Cawchy87
quote:
Originally posted by: rbV5
Isn't it pretty rare to damage a card by simply overclocking? I have to believe that modding a card has far more potential to actually damage a card. You break it, you buy it.



Yes... that's what people don't seem understand. Just increasing the speed doesn't hurt anything directly. Increasing the voltage can, and it getting hotter can.

*EDIT* On the other hand... putting an nVidia heatsink on an ATI card which doesn't make contact with the core will definately damage the card and is pure stupidity and should not be RMA'd.

agreed

Actually, power consumption for most things will increase with clock frequency, but increases roughly with the square of voltage increases.

The reason OCing without voltage increases is unlikely to break anything is that the part is likely to simply 'not function' at high overclocks, and therefore not get hot.

If you had a part with one weak link to act as a fuse, you could indeed break it just by OCing it.

In practise, it is indeed unlikely that this will happen without voltage increases or physical modification to the part. When I say I won't OC anything with a warranty, I consider that:

1. Getting a meaningful OC from most parts requires more voltage or modifications. Even a seemingly tiny voltage increase can have a very large effect on the power dissipation of the part, especially when coupled with clock speed increases.

2. Even if I get a good OC with no modifications, I'm still running the part out of spec. This even applies if I buy a 2500+ Barton, and a great cooler, and then run it as a 3200+. It uses more power, even at default voltage. I might consider buying one of these and doing this, and if it broke immediately when I tried to boot it as a 3200+ I might come to the conclusion that it was defective already, since this definitely shouldn't happen, but if I had been running it even for an hour or two before it broke, and even if the temperatures were nice and low, it's me that is SOL if it breaks, and that's how it should be. By purchasing a cheaper part, I'm also choosing to self-insure against product failure. AMD offers a part with a warranty at 3200+. It costs a lot more; I could buy two 2500+ Bartons and still be ahead money; if I'm forced to do so because OCing causes the chip to fail, it's my own damn fault!

Companies that support OCing under warranty are changing their warranties to meet customer needs and wants; there's nothing wrong with this, but those policies only apply to the parts they apply to. We as consumers should not feel we can re-write warranties at our leisure.

 

[Jeff7181]

Originally posted by: 3chordcharlie

Originally posted by: Cawchy87
quote:
Originally posted by: rbV5
Isn't it pretty rare to damage a card by simply overclocking? I have to believe that modding a card has far more potential to actually damage a card. You break it, you buy it.



Yes... that's what people don't seem understand. Just increasing the speed doesn't hurt anything directly. Increasing the voltage can, and it getting hotter can.

*EDIT* On the other hand... putting an nVidia heatsink on an ATI card which doesn't make contact with the core will definately damage the card and is pure stupidity and should not be RMA'd.

agreed

Actually, power consumption for most things will increase with clock frequency, but increases roughly with the square of voltage increases.

The reason OCing without voltage increases is unlikely to break anything is that the part is likely to simply 'not function' at high overclocks, and therefore not get hot.

If you had a part with one weak link to act as a fuse, you could indeed break it just by OCing it.

In practise, it is indeed unlikely that this will happen without voltage increases or physical modification to the part. When I say I won't OC anything with a warranty, I consider that:

1. Getting a meaningful OC from most parts requires more voltage or modifications. Even a seemingly tiny voltage increase can have a very large effect on the power dissipation of the part, especially when coupled with clock speed increases.

2. Even if I get a good OC with no modifications, I'm still running the part out of spec. This even applies if I buy a 2500+ Barton, and a great cooler, and then run it as a 3200+. It uses more power, even at default voltage. I might consider buying one of these and doing this, and if it broke immediately when I tried to boot it as a 3200+ I might come to the conclusion that it was defective already, since this definitely shouldn't happen, but if I had been running it even for an hour or two before it broke, and even if the temperatures were nice and low, it's me that is SOL if it breaks, and that's how it should be. By purchasing a cheaper part, I'm also choosing to self-insure against product failure. AMD offers a part with a warranty at 3200+. It costs a lot more; I could buy two 2500+ Bartons and still be ahead money; if I'm forced to do so because OCing causes the chip to fail, it's my own damn fault!

Companies that support OCing under warranty are changing their warranties to meet customer needs and wants; there's nothing wrong with this, but those policies only apply to the parts they apply to. We as consumers should not feel we can re-write warranties at our leisure.

The chances of that happening are slim to none... all the processors with "weak links" are weeded out during the burn-in process done by the manufacturer.

 

[3chordcharlie]

Originally posted by: Jeff7181
The chances of that happening are slim to none... all the processors with "weak links" are weeded out during the burn-in process done by the manufacturer.

I know that - and if I had $100 to spend on something I don't need, and a motherboard that supported 400fsb, I would gladly take the chance. But it would still be me taking a chance and not the company that makes or sells the product.

As I said, if it immediately and permanently failed on the first boot-up@400fsb, with the fan properly installed etc, I would consider that to be a defective part; and as you said, that's highly unlikely. Any other circumstances of failure and it clearly becomes my fault. I've burned out a chip before because the fan appeared to be installed correctly (it even stayed mounted when I turned the case right-side up). But it wasn't installed properly, and it was me who was out the $50, not anyone else. I've also destroyed a motherboard with a screwdriver due to a badly designed fan clip - again, my fault, I chose to buy the cheap fan, and that decision cost me $100 for a new motherboard!

 

[Jeff7181]

If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

 

[3chordcharlie]

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

Yeah, truly defective parts are pretty rare. Most things I've ever had to warranty-return were damaged in shipping. The odd thing has been wrecked by a buggy firmware update (not a failed update; bad releases that were later retracted).

 

[Fern]

Originally posted by: 3chordcharlie

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

Yeah, truly defective parts are pretty rare. Most things I've ever had to warranty-return were damaged in shipping. The odd thing has been wrecked by a buggy firmware update (not a failed update; bad releases that were later retracted).

Hmmm, some people may have better luck than others. I've gotten a defective mobo, a defective psu (half the 5volt 4 pin connectors were doa) and a defective HSF combo (fan was clattering upon startup). And a couple others I attribute to shipping (box was crushed as were the dimm retaining clips) Oh well, luck of the draw I guess.

 

[Jeff7181]

Originally posted by: Fern

Originally posted by: 3chordcharlie

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

Yeah, truly defective parts are pretty rare. Most things I've ever had to warranty-return were damaged in shipping. The odd thing has been wrecked by a buggy firmware update (not a failed update; bad releases that were later retracted).

Hmmm, some people may have better luck than others. I've gotten a defective mobo, a defective psu (half the 5volt 4 pin connectors were doa) and a defective HSF combo (fan was clattering upon startup). And a couple others I attribute to shipping (box was crushed as were the dimm retaining clips) Oh well, luck of the draw I guess.

No doubt defective parts make it out of the manufacturer... but it's a VERY small percentage considering how many good parts come out...

 

[ss284]

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

Unless its a dell 2001fp (I'm getting my third replacement tomorrow!) or an IBM 75gxp deathstar.

There are always exceptions.

And about the RMA'ing stuff, if the motherboard provides the option to up the fsb and the voltage out of the box, I have no problem RMAing it if it happens to die one day while overclocked. Physically modding the motherboard however, should instantly void the warranty.


-Steve

 

[Childs]

I would if I was just overclocking, but not something like a voltage mod. BFG sorta encourages overclocking, so I dont think its necessarily wrong. Replacing the cooling is a grey area. If it didnt work right after I replaced it I'd think I broke it and not RMA it, but if it dies months after I'd put the cooling back to stock and RMA it. If they challenge it then I probably wouldn't challenge it, depending on their analysis.

 

[jiffylube1024]

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

As a subsection of the 'damaged during shipping or improper handling before the item reaches the consumer' I'd add in 'mishandling by the OEM, which is probably surprisingly big. I've seen RAM, hard drives, and other sensitive parts mishandled by store clerks time and time again. Plus some display cases that are in direct sunlight aren't exactly great for most computer parts, either...

 

[DAPUNISHER]

Originally posted by: jiffylube1024


As a subsection of the 'damaged during shipping or improper handling before the item reaches the consumer' I'd add in 'mishandling by the OEM, which is probably surprisingly big. I've seen RAM, hard drives, and other sensitive parts mishandled by store clerks time and time again. Plus some display cases that are in direct sunlight aren't exactly great for most computer parts, either...

I was watching the tube for a little bit before going to bed the other night, and I saw that kook on shopNBC pushing his overpriced P4 systems. Well, they want to assure you they are a huge OEM and they show video of some of their facilities. You see the line with workers assembling the PCs, and this goober takes the ATAPI he is putting in the case and gives it a couple of good whacks with his hand to get it in the slot I don't know about you fellas, but when building PCs I most assuredly do not strike the components, potentially causing shock or physical damage to them. I'm certain other questionable practices are involved with the assembly and shipping of many OEM parts too. I've gotten boxed Intel CPUs with multiple bent pins! WTF???

 

[ZobarStyl]

I wouldn't return anything that performed reasonably at stock and failed under overclocking conditions: with video cards you should know which brands offer low quality products that are likely to fail and which are built with o/c'ing in mind. However, even brands that come with quality out-of-the-box cooling solutions should be o/c'ed carefully, like anything else. If you o/c something slowly and test conditions well, you shouldn't have any real problems because you'll 'see' the upper limit approach before you hit it and cause permanent damage. If you softmod/hardmod something it's your loss and RMA'ing would be almost criminal in my book; before you hardmod test your luck on old hardware and know your stuff before you ruin a 230 dollar card and if you softmod you knew it wasn't intended to run with that BIOS, what'd you expect?

 

[MysticX23]

to add on to this topic, i agree with most of you that its wrong to rma something that was oc'ed and died cuz of that...

...HOWEVER, to me it seems like ATI/NVIDIA are asking for it. I'm pretty sure they could simply lock up the oc'ing ability (like Intel's multiplier, etc) and there would be no more problems from cards being oc'ed.

 

[MysticX23]

I'm not sure about this part, but , it seems like that they "suggest" overclocking, or why else would they make it so that the core/mem speeds can change? If I were ATI or Nvidia, I'd just lock all the freq and I might lose some of the overclocking "enthusiasts", but i'm pretty sure that the percentage of RMA's would go down.

 

[MercenaryForHire]

No.

- M4H

 

[chsh1ca]

I find it interesting that there are 30 voices (at time of posting) answering Yes, however they have remained silent in this thread.

Originally posted by: DAPUNISHER
I was watching the tube for a little bit before going to bed the other night, and I saw that kook on shopNBC pushing his overpriced P4 systems. Well, they want to assure you they are a huge OEM and they show video of some of their facilities. You see the line with workers assembling the PCs, and this goober takes the ATAPI he is putting in the case and gives it a couple of good whacks with his hand to get it in the slot I don't know about you fellas, but when building PCs I most assuredly do not strike the components, potentially causing shock or physical damage to them. I'm certain other questionable practices are involved with the assembly and shipping of many OEM parts too. I've gotten boxed Intel CPUs with multiple bent pins! WTF???

Have you actually tried to render a motherboard nonfunctioning via the application of force (whether it be bending the motherboard, or pushing real hard on it with a screwdriver, etc)? It takes quite a bit of force to damage those boards -- even most PCI boards, let alone drives. Sure, most people can do it, but they are designed to survive such incidental contact. Yeah, it probably isn't the BEST idea to be hammering CD/DVD drives into place, but it probably isn't unique to that place.

Edit: Of all the DOA RMAs I've encountered, bent pins on OEM processors would take the cake in terms of both number and inconvenience. Three in a row caused me to stop doing business with one of my suppliers.

 

[MysticX23]

I have a feeling that those who want to say "yes" are afraid of getting banned, like the guy from the thread in the first post.

 

[Cawchy87]

Originally posted by: ss284

Originally posted by: Jeff7181
If I had to guess, I'd say 75% of items like this that are returned because they're damaged were damaged by the owner being careless or ignorant... another 24.5% are damaged somewhere during shipping or improper handling before they reach the consumer... and the other 0.5% are defective parts that were missed during quality testing by the manufacturer.

Unless its a dell 2001fp (I'm getting my third replacement tomorrow!) or an IBM 75gxp deathstar.

There are always exceptions.

And about the RMA'ing stuff, if the motherboard provides the option to up the fsb and the voltage out of the box, I have no problem RMAing it if it happens to die one day while overclocked. Physically modding the motherboard however, should instantly void the warranty.


-Steve

I am on my 4th dell monitor (crt tho) and it keeps flashing on and off. They won't send me a new one but "garentee" the next one works, sorry i had to get that off my chest.

 

[VIAN]

:disgust:Look at all these A-Holes that RMA cards that they broke. :|Stupid ass people can't OC and then make me pay for it.

 

[Schadenfroh]

Originally posted by: VIAN
:disgust:Look at all these A-Holes that RMA cards that they broke. :|Stupid ass people can't OC and then make me pay for it.

some arnt even broke, they RMA becuase they dont mod or overclock like they want.

 

[413xram]

No I would not. Plus when I really overclock a video card it is heavily modded with hardware adjustments. Such as: silicone, neoprene, bigger heatsink, conforming silicone, TEC, its kind of hard to get all that crap off anyway. It would look quite silly trying to RMA that...LOL That is why I have a BFG 6800 U watercooled at 470 MHz now. No heat or cooling problems just gaming. Sure I overclock it some..474 MHz...but there is really no need to get extreme. As long as I can game at 1600x1200 why get extreme? Probably in a year or so when this card is becoming obsolete I may bust out some extreme cooling, but right now I'm quite satisfied.))) Plus I'm really enjoying watching all these GTs overclocking without proper cooling. "Why is my card not stable?" LOFL..."Why do I show artifacts?") If I ran a GT and was going to overclock it I would be all over a TEC (maybe) with watercooling that would take care of GPU heat problems. I would definetly not overclock like I see people doing with stock cooling. Of course I do not have that problem

 

[Zebo]

I can't even be bothered to turn in free money in the form of rebates you think I'd waste my time for something I broke? Na/ Much better things to do like try out a new piece of equipment.

 

[Zebo]

Originally posted by: Jeff7181

Originally posted by: rbV5
Isn't it pretty rare to damage a card by simply overclocking? I have to believe that modding a card has far more potential to actually damage a card. You break it, you buy it.

Yes... that's what people don't seem understand. Just increasing the speed doesn't hurt anything directly. Increasing the voltage can, and it getting hotter can.

*EDIT* On the other hand... putting an nVidia heatsink on an ATI card which doesn't make contact with the core will definately damage the card and is pure stupidity and should not be RMA'd.

Sure it does. You're increasing current (aka one of the varibles to arrive at power) by increasing speed which increases heat output illustrated be the following equation:

Total Watts = Default Watts * (Overclocked Mhz \ Default Mhz) * (Overclocked Vcore \ Default Vcore)^2

Just that increasing voltage increases it more. Why do you think a 1.8 and 2.4 barton both running at same 1.65 vcore have very different temperature signatures?

 

[nick1985]

no. i say hang people who are caught doing it. hell, ill kick the chair out from under 'em.

 

[nRollo]

I have got to say how impressed I am by the three threads referenced here.

When the community as a whole starts condemning these losers for the stealing scum they are, everybody wins. (except doofus scum I suppose)

Kudos to all of the honest people with integrity.