Is my old PSU gonna be enough for an upgrade?

[dclive]

Originally posted by: Gillbot
All I will say is, jonnyGURU gave me similar upgrade my PSU asap advice when I built my C2D rig. He pleaded with me to upgrade but a few people said I;d be fine so I let it go and used my old PSU. It was a generic 430 watter and guess what, the system booted right up fine and ran for a few days before the PSU went POP! It took out a few other components with it though so why risk a shiny new build over an OLD PSU? It's not worth it, get a better one.

Interesting. What system did you have before this, and how many hundreds of people did an upgrade from an old system and never had problems?

In other words, sample size of one isn't that interesting. Sorry; no offense.

(To further top it off, newer systems, particularly the Core 2 Duo series, take LESS power than older systems, particularly the AMD64, AMD Athlon XP, and Intel P4 series.)

 

[MarcVenice]

No I'm calling bullshit because I don't believe that many people upgrade OEM pc's. Yes people might be asking on that dell forum or whatever, is that 1% out of all OEM buyers ? 10% ? Numbers, or it doesn't mean anything at all. You can't just 'assume' with an industry this big. Get some salesnumbers, should be doable I guess for the PSU and GPU industry, knowing there's only nvidia/ati.

 

[dclive]

Numbers of what, exactly?

Look at the micro part of this - take Best Buy. They sell lots of GPUs, but no (AFAIK) PSUs. Same with Circuit City.

That's your proof.

 

[jonnyGURU]

Originally posted by: dclive
Sounds logical enough - your PSU pulls 1000W from the wall, and you need at least an 800W PSU, or about 20% less than that.

Read my post. Don't skim it.

I had to really PUSH the build to get it to pull 1000W from the wall. WAY more pushing then Tom or Anand has ever done. Using the same methodology Tom or Anand use, I only pull 800W from the wall. Assuming 80% efficiency, that's only 640W.

THE MACHINE WON'T BOOT WITH LESS THAN 700W AND ISN'T STABLE WITH < 800W.

Apply this to a PC with lower demand: 320W DC during said tests. Will the PC even boot with a 350W? Will it even be stable with a 400W?

Originally posted by: dclive

Originally posted by: jonnyGURU
Now back ON topic... Yes, the OP's system with it's "mere" x1650 does not require much power at all. I can totally agree on that. But we shouldn't use "Kill-A-Watt-used-during-3DMark" number to drive that point home. The fact is, the OP has a cheap power supply. The OP has an OLD power supply. The OP's power supply likely was NEVER capable of 400W to begin with, yet you're telling him "you'll be fine."

He'll be fine with an Earth Watts 380W, sure. He'd be fine with a good FSP 350W even. No problem. But NOT the CompUSA 400W.

I find that highly unlikely given the millions of similar PSUs out and about today, the vast majority running perfectly fine. Fact is most people running 1650s and higher GPUs use cheap OEM PSUs and they're still working just fine.

The fact of the matter is, although the majority of them are "working just fine", a larger percentage of them than a decent quality, adequately sized PSU are failing withing a shorter period of time.

You seem to have the tunnel vision of "works now, works forever" without taking into consideration that running a PSU, especially a cheaply made one, is shortening it's lifespan exponentially.

Originally posted by: dclive
I see this as essentially a step-up of an OEM PSU. As we've already proven, an OEM PSU (Dell, Acer are 2 examples) at 300W and 350W are fully capable of running an 8800GTS/320 and Quad-core CPU, so to run a quad-core CPU and significantly less stuff and a significantly lower power GPU, with a PSU that's about as good and has more wattage per the maker, strikes me as a win-win scenario.

OMG! That's the problem right there. You actually believe that a CompUSA PSU is a "step up" from a Dell, HP, etc. PSU? Not even close. I would totally take a 300W Dell PSU over a 400W CompUSA PSU.

Originally posted by: dclive
Are you really saying you think everyone running a 1650 or another add-in GPU is going out to buy another PSU? C'mon!

I never said that. I'm saying someone with a shitty CompUSA that's going to be run dry and put away wet is probably in need of replacement.

 

[jonnyGURU]

Originally posted by: dclive

Originally posted by: MarcVenice
Maybe it's a bad test, not to interested in getting technical, like I said, there will be gpu and cpu bottlenecks, different games will draw different kinds of power.

Originally posted by: dclive
You think all those people buying 1650s at retail are also buying a PSU to go with it? C'mon! You are fully aware that the lion's share of machines out there are pre-built by the OEMs (Dell, etc.) and nobody touches them except for the few parts they update, right? I'd bet graphics cards outsell PSUs 10 to 1 at retail, if not by more.

Get some numbers dclive, because I'm calling bullshit. I'd be more inclined to actually believe that people who buy dells and other OEMs NEVER upgrade. Most of them don't know how. Dell's aren't easy to upgrade either. And you are talking 1650's here, what if we're talking 200$ gpu's instead? Don't you think THOSE people buy a new PSU ? It's all to easy to assume that people who think they need a better GPU don't also think need a better PSU, when the box clearly sais: 26a on the 12v rail required.

You're calling bullshit because you believe people upgrade PSUs like they upgrade GPUs? Oh come on! Until a few years ago you *couldn't* upgrade Dell PSUs (yes, there's an adapter, but most people didn't know where to get it...) - look at BestBuy and Circuit City sometime; they have massive varieties of graphics cards, but do they even sell PSUs? That's retail, and that's your proof....

Dells are trivial to upgrade the GPU - undo the case, slap in the GPU. Did you mean something else here?

For $200 GPUs (8800GT, currently), no, I don't think they all buy new PSUs. Take a look in the slickdeals Inspiron 530 (Quad core $675 with 24" LCD) thread sometime - lots of people asking, and 100% success rates with the stock PSU.

You're correct dclive. You're absolutely correct. Most people DO NOT upgrade their PSU's. And of those people, most of those people DO NOT have a problem. But the people that DO have a problem may not think that their problem is due to the PSU being inadequate or failing. They'll RMA a couple graphics cards before it gets through their skull if they're not lucky or unlucky enough to have the PSU die on them in the process of upgrading.

The point is, as Gillbot stated, why take the chance?

Will someone's stock PSU die when their system is upgraded? Probably not.

Will someone start experiencing hard to diagnose problems? Maybe, but not likely.

Will someone suspect a failing or inadequate PSU if they start having problems? Very likely no. In fact, the logical approach is to assume the bad part is the last part replaced.

The fact is, the CompUSA PSU is not a good unit. Certainly not as good as you think they are if you're putting them up against Delta and Bestec made Dell units for comparison. They're rated at peak, they're rated at room temp.... even Dell PSU's are rated for continuous output at operating temperature.

Is the PC going to burst into flames? No. Never said it would. Will it last a year? Who knows. Is it worth taking the chance? I think not.

 

[dclive]

Originally posted by: jonnyGURU
I had to really PUSH the build to get it to pull 1000W from the wall. WAY more pushing then Tom or Anand has ever done. Using the same methodology Tom or Anand use, I only pull 800W from the wall. Assuming 80% efficiency, that's only 640W.

THE MACHINE WON'T BOOT WITH LESS THAN 700W AND ISN'T STABLE WITH < 800W.

Apply this to a PC with lower demand: 320W DC during said tests. Will the PC even boot with a 350W? Will it even be stable with a 400W?

If it won't boot with less than X and isn't stable with less than Y, don't you think Tom and Anand would have found that? Seems like you've got the numbers down pat.

The fact of the matter is, although the majority of them are "working just fine", a larger percentage of them than a decent quality, adequately sized PSU are failing withing a shorter period of time.

You seem to have the tunnel vision of "works now, works forever" without taking into consideration that running a PSU, especially a cheaply made one, is shortening it's lifespan exponentially.

Interesting. On what do you base your belief? I've *never* upgraded a PSU, except one that I upgraded (a self-built machine) when eVGA tech support told me I needed it - and the reality was that when I went from a 6200 to an 8800, the ATITool or similar tool had a registry preference installed that caused the 8800 not to work (it was attempting to turn the 6200 into a 6600, and when the 6200 turned into an 8800, bad things happened...) - a waste of $80; putting in the original PSU worked just fine, too. If I had a cheaper PSU, on what do you base your belief it will fail more quickly? Is there a formula you use, or is this all a WAG?

Shortened lifespan - so shortened from 20 years to just 10? On what do you base your assertions? C'mon - these are the guys that overclock the heck out of their CPUs, and you're worried about running a PSU at UNDER its' specs?

OMG! That's the problem right there. You actually believe that a CompUSA PSU is a "step up" from a Dell, HP, etc. PSU? Not even close. I would totally take a 300W Dell PSU over a 400W CompUSA PSU.

That's fine. Either way, his PSUs going to sit there twiddlings its' thumbs.

I never said that. I'm saying someone with a shitty CompUSA that's going to be run dry and put away wet is probably in need of replacement.

Amazing how CompUSA was able to sell so many of them.... (30 seconds for someone to insert obligatory CompUSA comment here...)

 

[dclive]

Originally posted by: jonnyGURU
You're correct dclive. You're absolutely correct. Most people DO NOT upgrade their PSU's. And of those people, most of those people DO NOT have a problem. But the people that DO have a problem may not think that their problem is due to the PSU being inadequate or failing. They'll RMA a couple graphics cards before it gets through their skull if they're not lucky or unlucky enough to have the PSU die on them in the process of upgrading.
The point is, as Gillbot stated, why take the chance?
Will someone's stock PSU die when their system is upgraded? Probably not.
Will someone start experiencing hard to diagnose problems? Maybe, but not likely.
Will someone suspect a failing or inadequate PSU if they start having problems? Very likely no. In fact, the logical approach is to assume the bad part is the last part replaced.
The fact is, the CompUSA PSU is not a good unit. Certainly not as good as you think they are if you're putting them up against Delta and Bestec made Dell units for comparison. They're rated at peak, they're rated at room temp.... even Dell PSU's are rated for continuous output at operating temperature.
Is the PC going to burst into flames? No. Never said it would. Will it last a year? Who knows. Is it worth taking the chance? I think not.

Define "the chance" over, say, 2 years - 100% failure rate? 90% failure rate?

 

[jonnyGURU]

Originally posted by: dclive

Originally posted by: jonnyGURU
I had to really PUSH the build to get it to pull 1000W from the wall. WAY more pushing then Tom or Anand has ever done. Using the same methodology Tom or Anand use, I only pull 800W from the wall. Assuming 80% efficiency, that's only 640W.

THE MACHINE WON'T BOOT WITH LESS THAN 700W AND ISN'T STABLE WITH < 800W.

Apply this to a PC with lower demand: 320W DC during said tests. Will the PC even boot with a 350W? Will it even be stable with a 400W?

If it won't boot with less than X and isn't stable with less than Y, don't you think Tom and Anand would have found that? Seems like you've got the numbers down pat.

Why would they find that? Did they run these tests on 400W power supplies? No. They did not. Did they?

I have. Because it's part of QC when I was doing systems for a living and now that I work for a company that has to qualify minimum system requirements for graphics cards.

Originally posted by: dclive

Originally posted by: jonnyGURU
The fact of the matter is, although the majority of them are "working just fine", a larger percentage of them than a decent quality, adequately sized PSU are failing withing a shorter period of time.

You seem to have the tunnel vision of "works now, works forever" without taking into consideration that running a PSU, especially a cheaply made one, is shortening it's lifespan exponentially.

Interesting. On what do you base your belief? I've *never* upgraded a PSU, except one that I upgraded (a self-built machine) when eVGA tech support told me I needed it - and the reality was that when I went from a 6200 to an 8800, the ATITool or similar tool had a registry preference installed that caused the 8800 not to work (it was attempting to turn the 6200 into a 6600, and when the 6200 turned into an 8800, bad things happened...) - a waste of $80; putting in the original PSU worked just fine, too. If I had a cheaper PSU, on what do you base your belief it will fail more quickly? Is there a formula you use, or is this all a WAG?

I don't like pulling the "what I do or have done for a living" card, but if I have to... I have been a PC tech for well over a decade working for major e-tailers, system builders and even tier one distributors and half of my time spent at these businesses have been in troubleshooting, technical support, RMA and RTV. I know the numbers.

Originally posted by: dclive
Shortened lifespan - so shortened from 20 years to just 10? On what do you base your assertions? C'mon - these are the guys that overclock the heck out of their CPUs, and you're worried about running a PSU at UNDER its' specs?

Depends on the unit. But I know these numbers too.

MTBF is lowered exponentially at higher temperatures and greater loads. MOST units MTBF is rated at 25°C (room temperature) which is why you see MTBF as high as 10 years on really crappy units. The MTBF is also done at less than max capability, and by max I mean continuous, not peak. So you can easily turn a claimed "MTBF @ 25°C @ 80% load = 80K hours" into "MTBF @ 40°C @ 100% load = 10K hours" (a little over 1 year).

Originally posted by: dclive

Originally posted by: jonnyGURU
OMG! That's the problem right there. You actually believe that a CompUSA PSU is a "step up" from a Dell, HP, etc. PSU? Not even close. I would totally take a 300W Dell PSU over a 400W CompUSA PSU.

That's fine. Either way, his PSUs going to sit there twiddlings its' thumbs.

Originally posted by: jonnyGURU
I never said that. I'm saying someone with a shitty CompUSA that's going to be run dry and put away wet is probably in need of replacement.

Amazing how CompUSA was able to sell so many of them.... (30 seconds for someone to insert obligatory CompUSA comment here...)

Lame. There's lots of companies out there selling product with 10% RMA rates within one year's time period and they stay in business, etc. Fact is: It is (was) only CompUSA as a retail store that had to protect it's name. They didn't have to protect their name as a PSU company. If a company like OCZ or Corsair put out crap like that and even had HALF the RMA rate that CompUSA has (had) they would be out of business in three years.

 

[dclive]

Originally posted by: dclive

Originally posted by: jonnyGURU
Is the PC going to burst into flames? No. Never said it would. Will it last a year? Who knows. Is it worth taking the chance? I think not.

Define "the chance" over, say, 2 years - 100% failure rate? 90% failure rate?

So you think it might fail (that's a summary of your prior message) - so again, sure, fail at what rate? 100%? 90%? That's the question everyone wants to know. No one is shocked that there are differences in quality between units - what I think people would like to know is how to quantify those differences. Basically you suggest people should spend $50-$100 on the PSU upgrade, but don't quantify the risk. If you can't do that, no one can judge how critical this is.

(And the BestBuy/CircuitCity crowd, having not done PSU upgrades by and large when they upgrade GPUs, kinda suggests it's not critical.)

 

[dclive]

Originally posted by: jonnyGURU

Lame. There's lots of companies out there selling product with 10% RMA rates within one year's time period and they stay in business, etc. Fact is: It is (was) only CompUSA as a retail store that had to protect it's name. They didn't have to protect their name as a PSU company. If a company like OCZ or Corsair put out crap like that and even had HALF the RMA rate that CompUSA has (had) they would be out of business in three years.

What RMA rate did CompUSA have?

 

[jonnyGURU]

Originally posted by: dclive
Define "the chance" over, say, 2 years - 100% failure rate? 90% failure rate?

Doesn't have to be. Even 10 or 20% is unacceptable. And if the unit blows up on the secondary side and takes out some hardware with it... is that acceptable? Oh well? Better luck next time?

 

[dclive]

What is the failure rate? I keep asking. Is it 10%? 20%? 90%? What is it? I'd like to know so that I can try to quantify the risk. I've never had a PSU failure, so for me the risk so far is 0%, but I'd like to know what you have seen.

 

[jonnyGURU]

First, I'll have you know that the CompUSA PSU's are made by Leadman.

When I was working for the disty that would actually do fullfillment for CompUSA for builds, cases and PSU's up until about 2005, the return rate was about 20%.

Now mind you, return rate and failure rate are completely different. In my experience, failure rate is less than 50% of return rate even after a retail store "tech" deems a product "defective" (some guys behind the counter don't even realize that you can't just flip the siwtch on the back of a PSU to get it to turn on.) On the other hand, a lot of "techs" will only look at a PSU problem as "does it turn on? Yes. Ok... Not a PSU problem." The customer goes home, still has problems, gives up, buys a Dell. So you can balance that "defective" rate somewhere between 10 and 20%.

But also keep in mind that the warranty for these units is only one year. So beyond that, the store and the supplier aren't going to have any data. Since it's MORE LIKELY to have the product die in the field rather than DOA or within the first year, it's pretty safe to assume that your failure rate is considerably higher than 10 to 20% over the course of two or three years.

Sure, YOU'VE never had a PSU die. Guess what? I've never had a Ford Explorer roll over on me. Doesn't mean it doesn't happen. It's cool to base an opinion on personal experiences. That's where opinions come from. But I'm telling you from experience that it's not in the best interest of the customer to use a power supply like this to power any kind of quality components and certainly not for any extended period of time.

Like I said, 10 or 20% failure is unacceptable. It DOES mean that the odds are with you, but even a 5% RMA rate is unacceptable. And if a percentage of that 5% destroys one or two more components in it's wake it's even more so unacceptable. A quality part tends to only have a 2% failure rate and an even smaller percentage of incidental damage.

You can tell someone that banging your head against the wall isn't going to cause brain damage and in most cases it won't. Odds are you'll knock yourself out first. But if you bang long enough and hard enough, you can cause brain damage.

If the OP has enough sense to come to a forum and ask about his PSU, why not save him from being potentially one of the folks that has the PSU die on them or difficult to troubleshoot problems? They have a known bad power supply with unrealistic specifications that are prone to failure. I just don't see what sense it makes to send him on his way with the same PSU, telling him he's going to be "fine".

 

[dclive]

So you propose, regardless of all other data, to throw any PSU in the trash that does not meet these specifications (what are the specs you propose?) and buy a new PSU from a list of brands?

What about the obvious counterargument that the vast majority of the public doesn't do this, and yet their PCs work fine with updated GPUs?

If something has a 20% failure rate, I've got to believe there's going to be a major look into the issue by management; can you give me any more data on the 20% rate you listed? Of 100 sold, you literally mean that 20 would come back within one year? I've got to believe that would get someone's attention fairly quickly, if for no other reason than that it's trivial to track this data and very expensive to handle RMAs, since you need to have a tech test and you must have shipping ship the new part; even with unskilled labor that still adds up fairly quickly.

Were you one of dozens of people that must have been required to handle this return rate?

 

[Gillbot]

Originally posted by: dclive
So you propose, regardless of all other data, to throw any PSU in the trash that does not meet these specifications (what are the specs you propose?) and buy a new PSU from a list of brands?

What about the obvious counterargument that the vast majority of the public doesn't do this, and yet their PCs work fine with updated GPUs?

If something has a 20% failure rate, I've got to believe there's going to be a major look into the issue by management; can you give me any more data on the 20% rate you listed? Of 100 sold, you literally mean that 20 would come back within one year? I've got to believe that would get someone's attention fairly quickly, if for no other reason than that it's trivial to track this data and very expensive to handle RMAs, since you need to have a tech test and you must have shipping ship the new part; even with unskilled labor that still adds up fairly quickly.

Were you one of dozens of people that must have been required to handle this return rate?

I still say, if you plan on dumping all those funds into a new system, why cheap out over $50 more to keep an old possibly outdated PSU? Why not get a better spec'd one and be safe? Why argue over such a small amount and an insignificant point such as it MIGHT be ok?

 

[dclive]

Because nobody's saying it 'might' be OK. The strongest counterargument raised so far is the opposite - that there 'might' be a problem - and the numbers I'm seeing (20% failure rates!?) look ... unusual, to say the least.

Also bear in mind that for many people, a "new system" that they're "dumping money" into might be just a new motherboard and RAM. Are we really saying that every time you do a system upgrade you need to upgrade your PSU, or re-evaluate it if it has been working perfectly for months/years? That $50 PSU might increase the cost of the upgrade by 50% then. If it isn't necessary, why do it?

 

[jonnyGURU]

If I didn't know you any better I'd say you were trolling.

I'm not picking on CompUSA. I've worked for CompUSA, worked for a disty that fulfilled for CompUSA and really never had any complaints about CompUSA as a store or company despite seeing so many "good riddence" posts when they decided to finally go out of business.

And it's not like they went out of business because of crappy power supplies. 20% failure rate is something they, or any other retailer offering a generic solution, can afford. Seriously. Who's going to buy a CompUSA power supply? Someone that looks online for all of the "OMG my CompUSA power supply blew up and took my whole computer out with it!!!!"? Nope. Somebody that just needs a PSU in a pinch and doesn't want to spend more than $40. Cost of doing business? Who cares? They're marking the product up 400%! They could have a 50% failure and not care. If they sold a better product, they'd reduce their RMA's but make less money overall. You tell me someone in management didn't already think about that.

The fact remains that the PSU's are made by Leadman, aka Powmax, aka DiabloTek, etc. and to argue that they are "adequate" at the very least and there is no potential hazard in using one is like arguing that the sky is blue. It doesn't take too much gumshoe work to find posts or reviews of Leadman made product blowing up well before they even reached their advertised capability (there's two in this review from Oklahoma Wolf).

Leadman is almost non-existant these days because enough people are wise enough to know that it's a brand to stay away from. Again, if there's a DAMN GOOD REASON to believe that this PSU will not be ok, albiet not the most likely scenario, why in the WORLD would you tell someone they'd be "fine" based on the fact that YOU have had better luck with better power supplies?

 

[dclive]

Again, the strongest we can say is there "might" be a problem, and you're willing to discount the fact that the PSU has been just fine until now. So it isn't that you don't think the PSU can handle the wattage/stress of running the components, you just think the PSU brand is bad, so it should always be replaced. Furthermore, if it isn't replaced, you think the failure rate of the PSU might be somewhere from 10 to 20 percent over a time period of a year or two.

Does that about sum it up?

 

[jonnyGURU]

Originally posted by: dclive
Again, the strongest we can say is there "might" be a problem, and you're willing to discount the fact that the PSU has been just fine until now. So it isn't that you don't think the PSU can handle the wattage/stress of running the components, you just think the PSU brand is bad, so it should always be replaced. Furthermore, if it isn't replaced, you think the failure rate of the PSU might be somewhere from 10 to 20 percent over a time period of a year or two.

Does that about sum it up?

Let me ask YOU something....

Do you still believe that this guy actually has a 400W power supply? In other words: A power supply that can do anywhere NEAR 400W on a good day because that's what the label says it can do?

Given the fact that most people that "test" these PSU's aren't able to load them up to even 50% of their "advertised capability" without having them blow up, I think it's pretty safe to say that this isn't a 400W power supply.

You DO realize that not all power supplies are created equal, right? One company's definition of "400W" can be completely different than another.

And I still think you ignore the fact that components do degrade over time is "stressed". Certainly a PSU of this caliber is "stressed" the moment it's used for anything more than a PC with on-board video (which would work with a 200W power supply just fine).

Again, even if the odds are AGAINST the unit blowing up, why throw caution to the wind?

 

[dclive]

If I had a report with perfect accuracy that had a list of the most-sold PSUs and their reliability - with specific real world test parameters, I'd feel a lot better about changing how I worked and what I bought.

Barring that, and *given that the average Joe User buys a GPU from his local retail store and everything works just fine* (which everyone here seems to be completely discounting!), I still have my doubts. So far even the strongest-minded here (that would be you, JG! <g>) have simply said it's slightly (10-20%) more likely to have a problem (but I haven't seen return rates for "good" brand PSUs, so I'm not sure what we're comparing against - 10% vs. 5%, for example?). We don't know what scenarios it's more likely to have a problem, nor do we know if (assuming the PSUs been working for months/years) it's worthwhile to buy a new PSU when you've got what appears to be a known-good off-brand PSU.

 

[dclive]

Originally posted by: jonnyGURU
Again, even if the odds are AGAINST the unit blowing up, why throw caution to the wind?

For the same reason we don't put bulletproof glass in our car windows, the same reason every single person doesn't buy flood insurance for their house, and the same reason we don't buy loss insurance for every single thing we buy - because the chances of something happening, in the grand scheme of things, is too low to worry about.

(Or to put it another way, because I'd rather put money into my 401(k) rather than into computers.)

 

[jonnyGURU]

Originally posted by: dclive
Barring that, and *given that the average Joe User buys a GPU from his local retail store and everything works just fine* (which everyone here seems to be completely discounting!), I still have my doubts.

I didn't discount that. You didn't read my post. I had said that MOST PEOPLE with a Dell, HP.. even an eMachine, for goodness sake, has a better PSU than what these house brand CompUSA, BestBuy, etc. units are.

If average Joe was buying a graphics card AND a cheap PSU when they walked into CompUSA, I think we'd see completely different results.

Originally posted by: dclive
So far even the strongest-minded here (that would be you, JG! <g>) have simply said it's slightly (10-20%) more likely to have a problem (but I haven't seen return rates for "good" brand PSUs, so I'm not sure what we're comparing against - 10% vs. 5%, for example?).

I did give you this. It's typically less than 2%. Even PSU's that you hear about "dying all the time" like the Antec's with swollen caps and the Enermax's with the exploding primary rectifiers STILL only fail at a rate of 5% over a couple years time.

And true, not everyone buys flood insurance nor do they have to. But does that mean it's smart to not buy flood insurance when you live in a flood plain? Does it not cost more to replace everything damaged in a flood than what the insurance would've cost you even in the long run?

OP's in a flood plain. Good to buy some flood insurance.

 

[Gillbot]

To me this is like arguing about Pep Boys' tires on a $100,000 supercar. It's not THAT much more to upgrade the PSU to be 100% safe. I've seen the "i've used my old PSU and it took out xxx on my new rig" posts way too many times for me not to recommend FOR an upgrade to be safe.

 

[superstition]

Dump it. The risk to your more expensive components isn't worth it. There are cheap Seasonic PSUs on NewEgg that are quiet and reliable.

 

[s44]

Originally posted by: dclive
Numbers of what, exactly?

Look at the micro part of this - take Best Buy. They sell lots of GPUs, but no (AFAIK) PSUs. Same with Circuit City.

That's your proof.

Totally wrong, like everything else you've written in this thread.