Beware of Exploding Gigabyte PSU's being Dumped by Newegg in Forced Bundles

Schro

Supreme [H]ardness
Joined
Mar 15, 2002
Messages
7,803
for psu's it is within spec to safely deal with the tested out-of-spec use

Where in the ATX12v/EPS spec is this specified? OPP is a feature that they decided to implement and is not part of the spec.


This is only an impossible position if the company stubbornly persists in its own righteous we did nothing wrong stance.
They directly addressed the GN complaint about OPP being set too high. In other news, GN didn't engineer the PSU nor were they aware of the design goals for the product or what was originally contemplated when setting the OPP where it was.


Guess which one does the least harm to their reputation regardless of the actual quality of the product in question.
I'm not sure either choice in response would yield a material difference in reputation harm. Recalling is admitting that they are wrong and then GN will take a victory lap on a regular basis reminding people how their work got the recall to happen.
 

emphy

Limp Gawd
Joined
Aug 31, 2016
Messages
298
I'm not sure either choice in response would yield a material difference in reputation harm. Recalling is admitting that they are wrong and then GN will take a victory lap on a regular basis reminding people how their work got the recall to happen.
So what? Gamersnexus' work only got so much "hype" due to gigabyte refusing to acknowledge the problems before publication. This despite the rather generous amount of time they had and numerous reports of the psu's failing before gn ever getting involved in the matter. A simple "wtf, those psu's should not do that! We'll recall immediately." would have gathered rather more sympathy than what they did now, which provided extra material for videos by making gigabyte look like jackasses who care more about profits and their own ego than they do about their user's equipment.

As I mentioned before, shit happens during manufacturing. Companies which refuse to acknowledge this tend to get it flung back at them, companies which address the problems much less. I find it rather surprising that I need to point out that a no-cow-droppings solution is appreciated more than a half-hearted one which pays significant attention to claiming that there was nothing wrong in the first place.
 
Last edited:

thecold

Gawd
Joined
Nov 12, 2017
Messages
988
You're making a different point though - you both agree that it's not a good power supply. What no one else seems to see is that the test method used to cause the fireworks isn't a very good test method.


I don't agree with you, but even if I did. As Gamers Nexus pointed out, this was just the tip of the iceburg. If I remember, the good psu that was only loaded 60 percent didn't even go into op, and it still died.

Also I've been looking at the comments periodically, these 2 are just bad psu's.

1629343707340.png


1629343760108.png
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
This is only an impossible position if the company stubbornly persists in its own righteous we did nothing wrong stance. Even if one accepts your premise that the testing methodology is wrong (which I don't; for psu's it is within spec to safely deal with the tested out-of-spec situation, safely being the operative word here), it is much better to take the hit and just recall those psu's instead of what would essentially be feeding the alleged hype. That way they would have turned a "gigabyte is risking our equipment for their profits"-perception into a "wow, gigabyte make damn sure that they don't damage our equipment" one. Guess which one does the least harm to their reputation, regardless of the actual quality of the product in question.

No, Schro is right. There is no specification for the "test" they applied and there is no way to design a unit for every possible scenario. When/if I do an OPP test the load tester increases the draw in set steps that are defined under StepI for a period of Step T until it reaches OPP End I in the programming rinse and repeat until it reaches 1560W or the unit blows up. That, however, is not found anywhere in the specification and it is in steps that plateau and hold before moving to the next that I define. So, I could make them impossibly short or impossibly long to get the results out of the test thart I WANT. Not what should actually happen. I could kill every single power supply I test if I wanted to.
 

Mr Evil

Weaksauce
Joined
Jul 11, 2015
Messages
92
Where in the ATX12v/EPS spec is this specified? OPP is a feature that they decided to implement and is not part of the spec.
OPP may not be part of the spec, but the spec does include various protections which are obviously supposed to provide protection against this sort of failure. If Gigabyte add any more protections beyond what are required by the ATX spec, then it's reasonable to expect them to actually work.

The ATX spect also contains the following, which is amusing in light of some of the clips in the GN video:
3.1.5. Catastrophic Failure Protection
Should a component failure occur, the power supply should not exhibit any of the following:
• Flame
• Excessive smoke
• Charred PCB
• Fused PCB conductor
• Startling noise
• Emission of molten material
 

FearTheCow

Supreme [H]ardness
Joined
May 2, 2006
Messages
5,826
I think their PSU's are hit or miss, depending on who they are rebranding this week, but I do agree, they stand by their stuff and have excellent customer service.

I remember that thread on here a few years back with some guy who had a Corsair AIO water cooler leak all over his computer. They essentially replaced all of his parts in an unusually good show of customer service.

I don't forget stuff like that.
That was probably me, i was a broke college student at the time,, they sent a check covering retail cost of all my components.

I have spent at least 3 to 4 times the amount they sent me at this point and will continue to do business with them.

On the other hand, I bought a gigabyte board to replace what the aio damaged. That board had what looked like a capacitor explode causing a flash and burn marks. I sent the board in for RMA, gigabyte said it still worked so refused to warranty it. I refuse to buy gigabyte products now.

Moral of my story, I am not surprised that gigabyte is refusing to do anything about these power supplies.
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
OPP may not be part of the spec, but the spec does include various protections which are obviously supposed to provide protection against this sort of failure. If Gigabyte add any more protections beyond what are required by the ATX spec, then it's reasonable to expect them to actually work.

The ATX spect also contains the following, which is amusing in light of some of the clips in the GN video:

Which one? Because the one you are going to cite is violated by almost 99% of PSU on the market.
 

Mr Evil

Weaksauce
Joined
Jul 11, 2015
Messages
92
Which one? Because the one you are going to cite is violated by almost 99% of PSU on the market.
I wasn't going to cite any particular one, but if you want me to state a specific output protection that is supposed to prevent failure, then short-circuit protection seems like a good example. It's something that has saved a few of my own power supplies over the years from accidents and "experiments" on my part. I would be disappointed by any PSU that didn't recover after a short.
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
I wasn't going to cite any particular one, but if you want me to state a specific output protection that is supposed to prevent failure, then short-circuit protection seems like a good example. It's something that has saved a few of my own power supplies over the years from accidents and "experiments" on my part. I would be disappointed by any PSU that didn't recover after a short.

Except this is not a short circuit or what was "tested". So that has nothing to do with the issue.
 

Ebernanut

[H]ard|Gawd
Joined
Dec 15, 2010
Messages
1,557
If anything, they should have done the standard load tests prior to running it out of specification. It's really that simple - don't do something that will damage the unit before starting your review procedures.


What was there to do though? GN harped on OPP, but reducing OPP on these units does not make them a better or safer unit. I wouldn't be shocked (oh, now that's punny) if GN's test methodology was applied to the revised units and got the same results.



They were also put in an impossible position due to the test methodology and boisterous presentation. Their response said that no one should be surprised that a PSU would fail after the manner of OPP testing and offered to lower the OPP of your unit in the event you were persuaded by GN that it was the root cause of the PSU not being all that great (even though it's not). There's literally no response that would satisfy the keyboard warriors other than them offering an exchange for 3 new Titanium Seasonic psus for every 750/850GM turned in.
They did do regular testing and had at least one failure there(as has been mentioned a couple times now) and the reason they did extreme testing was to figure out why so many end users were having failures. Also if you had actually watched the video you're criticizing you'd know that they actually pointed out that the OPP was most likely not the cause of the failures but something that made it easier to trigger the fault, again many of these units including one of their test units died under normal operating conditions.

GN gave them 5 months heads up before saying anything publicly and the other person they mentioned(I forget who it was) had been in contact with them last november and gigabyte stonewalled them and ignored the situation until it blew up after GN posted the video. Gigabyte has no one but themself to blame for their "impossible position".
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
They did do regular testing and had at least one failure there(as has been mentioned a couple times now) and the reason they did extreme testing was to figure out why so many end users were having failures. Also if you had actually watched the video you're criticizing you'd know that they actually pointed out that the OPP was most likely not the cause of the failures but something that made it easier to trigger the fault, again many of these units including one of their test units died under normal operating conditions.

GN gave them 5 months heads up before saying anything publicly and the other person they mentioned(I forget who it was) had been in contact with them last november and gigabyte stonewalled them and ignored the situation until it blew up after GN posted the video. Gigabyte has no one but themself to blame for their "impossible position".

That doesn't point to a design issue but a component issue, likely linked to one or a few batches of that component.
 

Ebernanut

[H]ard|Gawd
Joined
Dec 15, 2010
Messages
1,557
That doesn't point to a design issue but a component issue, likely linked to one or a few batches of that component.
They seemed to think it could be either at this point but that's above my paygrade. However the cause doesn't really matter to me as much as the way they handled the whole situation, they could have done a quiet recall when they first heard about it or at least stop shipping units with the faulty part if that's the issue but instead they did nothing until it blew up on them months later(they knew in november and were shipping affected units until at least may).
 

chrcoluk

[H]ard|Gawd
Joined
Jul 7, 2008
Messages
1,260
So what? Gamersnexus' work only got so much "hype" due to gigabyte refusing to acknowledge the problems before publication. This despite the rather generous amount of time they had and numerous reports of the psu's failing before gn ever getting involved in the matter. A simple "wtf, those psu's should not do that! We'll recall immediately." would have gathered rather more sympathy than what they did now, which provided extra material for videos by making gigabyte look like jackasses who care more about profits and their own ego than they do about their user's equipment.

As I mentioned before, shit happens during manufacturing. Companies which refuse to acknowledge this tend to get it flung back at them, companies which address the problems much less. I find it rather surprising that I need to point out that a no-cow-droppings solution is appreciated more than a half-hearted one which pays significant attention to claiming that there was nothing wrong in the first place.

Pretty much this, no company is immune to a bad run of products, and its how they react that makes me judge them.
 

cybereality

[H]F Junkie
Joined
Mar 22, 2008
Messages
8,490
Full stop. Any product should not explode. If it does, under any condition, then it is faulty and defective. We shouldn't make excuses for this.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
19,785
I'm not sure either choice in response would yield a material difference in reputation harm. Recalling is admitting that they are wrong and then GN will take a victory lap on a regular basis reminding people how their work got the recall to happen.

Gigabyte had plenty of time to recall before this became a big issue. They could have done it last year when it was first reported, or in the FIVE MONTHS between GN first contacting them about their results and the video being published. You know what would have happened then? Nothing. It would have been a minor blip on the radar. A company making a mistake and correcting it when they were notified of how bad it was. By now, no one would even have remembered it happened. Instead they ignored it until they were forced to respond and then responded in a way that does nothing but actually hurt their reputation. Even if they had done a recall right after the video was released they still could have saved face. GN might have taken a few off-hand shots at them going forward (not entirely undeserved ones either) but it would all blow over relatively quickly and within a few months no one would even think about it anymore. There is no “impossible situation” here for Gigabyte, just a company with it’s head too far up it’s own ass.
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
Gigabyte had plenty of time to recall before this became a big issue. They could have done it last year when it was first reported, or in the FIVE MONTHS between GN first contacting them about their results and the video being published. You know what would have happened then? Nothing. It would have been a minor blip on the radar. A company making a mistake and correcting it when they were notified of how bad it was. By now, no one would even have remembered it happened. Instead they ignored it until they were forced to respond and then responded in a way that does nothing but actually hurt their reputation. Even if they had done a recall right after the video was released they still could have saved face. GN might have taken a few off-hand shots at them going forward (not entirely undeserved ones either) but it would all blow over relatively quickly and within a few months no one would even think about it anymore. There is no “impossible situation” here for Gigabyte, just a company with it’s head too far up it’s own ass.

That also, would have required them to identify the problem exactly, Given that it does not sound like a design issue but an issue with a batch of components that kind of tracking is way harder, especially when it is a peripheral business and not part of your core business market. Not to mention the issue is sporadic.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
19,785
That also, would have required them to identify the problem exactly, Given that it does not sound like a design issue but an issue with a batch of componnents that kind of tracking is why harder, especially when it is a perfipheral business and not part of your core business market. Not to mention the issue is sporadic.

Which is something that Gigabyte could communicate instead of insisting that their units are perfect and that any test saying otherwise is invalid. At any point between now and before this blew up they could have told GN they were looking into the issue and working with the OEM to figure out why so many units were failing.
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
Which is something that Gigabyte could communicate instead of insisting that their units are perfect and that any test saying otherwise is invalid. At any point between now and before this blew up they could have told GN they were looking into the issue and working with the OEM to figure out why so many units were failing.

<shrug> And they may have. I have told vendors about faults over the years about all kinds of actual spec complaint stuff and some of them are better than others about responding. Some have dedicated departments to this and some have one guy or no guys who montor this......lots of this stuff is blind rebrand. Telling vendors has resulted in everything from a thank you and respin of a product to people threatening to sue me. But, the thing is, none of those things went Youtube holier than thou. The going Youtube holier than thou is a sure sign of someone trying to turn a buck on a probably not as big an issue as they make it out to be because they don't make as much on the Youtube if they are like "hey this might be an issue we are looking" versus "holy smoking balls batman". So, yeah, sure, you guys get all pitchforked up about it without understanding what it is and they get lots of sensational clicks that gives them money on their YouTube advertising views. Gee, what is the reason for doing this....hmmmmm..... So, why be really honest about an issue when being sensational gives you more money? Think about it.
 

cybereality

[H]F Junkie
Joined
Mar 22, 2008
Messages
8,490
I'm not sure I agree with that premise. I watched the videos, there are sparks flying it *is* "holy smoking balls batman" if I ever saw it.

Sure, GamersNexus might have sensationalized the issue for the YouTube audience, but there is a real issue here. I also went and read the first page of Newegg reviews to see for myself and the thing is a piece of shit.

If anything, it is a PR failure, maybe we can agree to that. If Gigabyte was responsible, they could have stopped the video from ever being produced and (if that was not possible) at least came up with a better story or response. They did neither.
 

[Spectre]

[H] Admin
Staff member
Joined
Aug 29, 2004
Messages
18,090
I'm not sure I agree with that premise. I watched the videos, there are sparks flying it *is* "holy smoking balls batman" if I ever saw it.

Sure, GamersNexus might have sensationalized the issue for the YouTube audience, but there is a real issue here. I also went and read the first page of Newegg reviews to see for myself and the thing is a piece of shit.

If anything, it is a PR failure, maybe we can agree to that. If Gigabyte was responsible, they could have stopped the video from ever being produced and (if that was not possible) at least came up with a better story or response. They did neither.

Yes, and when I reviewed it the unit it was not good. However, to get components to do what you say requires other integration issues at the OEM that Gigabyte could not control (you can argue they should have QC's it but not every vendor has the ability). Now, that said, and when I reviewed it i told people it was not a good unit and to avoid it. So, it is all just sensationalism when you can get a few to do what you want. It is like the old Ford/Firestone issue.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
19,785
Yes, and when I reviewed it the unit it was not good. However, to get components to do what you say requires other integration issues at the OEM that Gigabyte could not control (you can argue they should have QC's it but not every vendor has the ability). Now, that said, and when I reviewed it i told people it was not a good unit and to avoid it. So, it is all just sensationalism when you can get a few to do what you want. It is like the old Ford/Firestone issue.

If Gigabyte is unwilling to put in the effort to QC what they purchase then maybe they shouldn't be selling PSUs in the first place.
 

cybereality

[H]F Junkie
Joined
Mar 22, 2008
Messages
8,490
Yes, I can understand. It is not Gigabytes' core-business, they contract some low-quality OEM and slap their brand on it. I get it.

What I'm saying, is it doesn't seem worth it. They can't be making all that much money on a cheap part, and the reputation loss is far greater than the margin on this piece of junk.

But whatever. I am happy with my Gigabyte motherboard and the new Gigabyte monitor I got. I don't think the company is done, I'm just saying that they made a bad decision.
 

jfreund

[H]ard|Gawd
Joined
Sep 3, 2006
Messages
1,271
Don't forget Gigabyte and Newegg have been pushing these in shuffle bundles for, how long now?

"You want a video card? Take this ticking time bomb, too."

Gigabyte has known this line is crap. Techpowerup had their review unit pop and fizzle 10 months ago and Gigabyte's response was "We 'tested' some and it didn't happen here". Gigabyte didn't care enough to look at the failed review sample to diagnose the cause.

I'm curious what percentage of units sold were in bundles to customers that didn't even want a Gigabyte power supply.
 

Schro

Supreme [H]ardness
Joined
Mar 15, 2002
Messages
7,803
Gigabyte didn't care enough to look at the failed review sample to diagnose the cause.
It's not in the reviewer's interests to provide the manufacturer access to the failed sample as those are helpful for the reviewer in the event of a lawsuit. It becomes difficult to defend yourself in court if your evidence of the facts gets disappeared.
 

Mr Evil

Weaksauce
Joined
Jul 11, 2015
Messages
92
Except this is not a short circuit or what was "tested". So that has nothing to do with the issue.
You're the one that sked me the question.

Anyway, It's relevant in that the reason the ATX spec spends quite a lot of time describing protections is that PSUs are supposed to be able to safely cope with common fault conditions. This isn't the 20th century any more where electronics were dumb, we now quite reasonably expect things to not explode even when used incorrectly.
 

Nafensoriel

Limp Gawd
Joined
Nov 23, 2015
Messages
382
You're the one that sked me the question.

Anyway, It's relevant in that the reason the ATX spec spends quite a lot of time describing protections is that PSUs are supposed to be able to safely cope with common fault conditions. This isn't the 20th century any more where electronics were dumb, we now quite reasonably expect things to not explode even when used incorrectly.
Bahahahahaha
This is so beyond untrue.

"Smart" electronics just fail more creatively(and usually more often).

The main failure points are decided by physics and power delivery has always been a physics problem. No power supply is immune to all failures. All have faults and magic bbs that can toast them.
I don't really care about digital jesus or gigabyte in general but in the world of scientific equipment and industrial equipment power supplies are far more serious a business than the PC market will ever hope to achieve and even in that market "Weird Sh*t Happens".
 

jfreund

[H]ard|Gawd
Joined
Sep 3, 2006
Messages
1,271
It's not in the reviewer's interests to provide the manufacturer access to the failed sample as those are helpful for the reviewer in the event of a lawsuit. It becomes difficult to defend yourself in court if your evidence of the facts gets disappeared.
That's a completely new take to me, and not the stance of the reviewer that had the sample fail.

https://www.techpowerup.com/285612/...byte-statement-on-gp-p750-850-gm-design-flaws

"They didn't ask for my bad sample back for failure analysis, which is the typical procedure"
 

notarat

2[H]4U
Joined
Mar 28, 2010
Messages
2,348
It's not in the reviewer's interests to provide the manufacturer access to the failed sample as those are helpful for the reviewer in the event of a lawsuit. It becomes difficult to defend yourself in court if your evidence of the facts gets disappeared.

From jfreund's link:

Aris response: I reported the issues I faced with the GP-P750GM power supply in late October 2020, before I post its review on TechPowerUp and the video review on the Hardware Busters YT channel, and GIGABYTE responded that its engineers tested five units and found no problems. They didn't ask for my bad sample back for failure analysis, which is the typical procedure, and they didn't offer a second sample to continue the review. I have kept all of my correspondence with GIGABYTE's respective team, in case there is any doubt.

The response seems to clearly indicate the reviewer expected Gigabyte to request that he send them the bad unit for testing. He further states that is the typical procedure when he encounters a part that dies, and that Gigabyte did not follow the typical procedure.

They made a crap product and are handling the issue in the same crappy manner.
 

GoodBoy

2[H]4U
Joined
Nov 29, 2004
Messages
2,187
You're making a different point though - you both agree that it's not a good power supply. What no one else seems to see is that the test method used to cause the fireworks isn't a very good test method.
It's industry standard testing equipment. Gigabyte is trying to place blame on anything except their crappy product.
If anything, they should have done the standard load tests prior to running it out of specification. It's really that simple - don't do something that will damage the unit before starting your review procedures.


What was there to do though? GN harped on OPP, but reducing OPP on these units does not make them a better or safer unit. I wouldn't be shocked (oh, now that's punny) if GN's test methodology was applied to the revised units and got the same results.



They were also put in an impossible position due to the test methodology and boisterous presentation. Their response said that no one should be surprised that a PSU would fail after the manner of OPP testing and offered to lower the OPP of your unit in the event you were persuaded by GN that it was the root cause of the PSU not being all that great (even though it's not). There's literally no response that would satisfy the keyboard warriors other than them offering an exchange for 3 new Titanium Seasonic psus for every 750/850GM turned in.
All of that is Gigabytes own fault - they designed, or approved for oem resale, an absolute piece of shit for a power supply.

Testing: the design allows it to run out of spec - the testing was fine. You ignore all of the failures people had that did nothing in regards to 'testing'.

The units' OPP: just another example of complete failure in the design.

The 'boisterous' presentation were 100% justified and necessary, to point out the utter disregard Gigabyte had to the large amounts of failures, and their repeated "just brush it under the rug" mentality. They needed called out on it, and called out without pulling any punches.

If you want to keep buying their products, go ahead. You aren't convincing any of the rest of us to do so.
 

DukenukemX

Supreme [H]ardness
Joined
Jan 30, 2005
Messages
5,875
They were also put in an impossible position due to the test methodology and boisterous presentation.
Them running games was an unreasonable test?
There's literally no response that would satisfy the keyboard warriors other than them offering an exchange for 3 new Titanium Seasonic psus for every 750/850GM turned in.
If Gigabyte isn't going to recall and refund money then a Season product of equal power rating would be fine. Realistically Gigabyte should just recall these products. Right now it's a bunch of upset keyboard warriors but next month it could be a similar situation where Firestone was selling faulty tires and killing people. These things explode and can start a fire or do something worse. Stop working for Gigabyte already.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
19,785
I did not say that. I said running the psu out of spec and then testing is unreasonable.

Not at all. A properly designed PSU will not have a catastrophic failure under those conditions. This has been standard procedure for PSU testing pretty much since anyone started to take PSU testing seriously. It's only a problem when there is a critical fault with the unit itself, whether a design flaw or a faulty part in the unit.

A power supply with protections in place should basically never fail the way these units did.
 

DukenukemX

Supreme [H]ardness
Joined
Jan 30, 2005
Messages
5,875
I did not say that. I said running the psu out of spec and then testing is unreasonable.
The PSU should ideally have protection against this. It isn't unreasonable to stress a PSU and then use it to see if everything is fine. Protection should kick in and prevent the PSU from going beyond spec. Considering how many bad NewEgg reviews there are, it's reasonable to assume that normal use of these PSU's is going beyond spec.
They haven't sent any money my way yet and I don't have time to work for free.
Then why are you defending Gigabyte? They're clearly in the wrong here and haven't taken reasonable actions, which should be to recall these products. These PSU's are a fire hazard waiting to get much worse. Instead Gigabyte wants buyers to contact them for a replacement which is not going to be most people who own these products, and they know that. All it takes is for someone to buy a RTX 3090 and use it with these PSU's and failure is inevitable.
 
Top