Speedrun Gigabyte Power Supply Explosion: Biggest Failure Yet (GP-P750GM)
Vložit
- čas přidán 27. 05. 2024
- Gigabyte kept claiming that media testing was over "extended time periods," so we decided to speedrun the explosion of one of its GP-P750GM PSUs to prove a point.
Sponsor: Buy Corsair's 5000D Airflow Case on Amazon (geni.us/cnVP60)
For context, here's how the PSU is supposed to work: • Gigabyte, This is How ...
THERE ARE ZERO CUTS IN THIS VIDEO BEFORE THE PSU FAILS. This is a single-take, one-shot speedrun to challenge Gigabyte's statements that third-party media testing was over "extended" periods of time, as Gigabyte tried to paint testing in the light of being ridiculous. The problem remains the same: Aris of Hardware Busters (Cybenetics) thinks that the MOSFETs are mistimed, and we agree. They're also not particularly good quality, and tend to go boom and hit runaway thermal scenarios (if not design faults first). OPP trips or not, we believe the Gigabyte GP-P750GM is poorly designed. The GP-P850GM has also had similar failures in our testing.
Support GN in its testing endeavors by getting something useful in return! You can grab a GN Wireframe Desk-Sized Mouse Mat, a Red & Black GPU component mouse pad, PC building Modmat work surfaces, and more: store.gamersnexus.net/
Watch Part 1 (Gigabyte PSU Fires): • Exploding Power Suppli...
Part 2 (Response to Gigabyte): • Gigabyte Twists Truth ...
Check out Aris' response here! • How to Get Away With A...
Get behind-the-scenes videos about other PSU testing as hosted by Patrick Stone: / gamersnexus
RELATED PRODUCTS
EVGA Super Nova 750 G5 on Amazon is actually good: geni.us/73O0h
Seasonic Focus GX-750 on Amazon is also actually good: geni.us/JNVMaI
TIMESTAMPS
00:00 - Uncut Speedrun, 2 Players
04:33 - Gigabyte Re-framing the Story
06:07 - Going to 60% Load
07:00 - Explaining the Test Interface
08:22 - 120% Load & First OPP Ramp
08:32 - Giving Gigabyte An Advantage in the Speedrun
09:22 - First Shut Down (OPP Trigger)
13:25 - Lab Testing is More Standardized Than Speedruns
14:00 - High OPP is NOT A GOOD THING!
16:52 - KABOOM - Headphone Warning!
17:25 - Stone Needs New Pants
20:31 - Explaining the Problem & Testing Approach
21:40 - "DC Electronic Load" Isn't "Not Real"
23:00 - A Computer is Potentially Even More Stressful!
23:30 - MOSFET Timings Might Be the Problem
25:50 - Gigabyte Can Do Better
** Please like, comment, and subscribe for more! **
Links to Amazon and Newegg are typically monetized on our channel (affiliate links) and may return a commission of sales to us from the retailer. This is unrelated to the product manufacturer. Any advertisements or sponsorships are disclosed within the video ("this video is brought to you by") and above the fold in the description. We do not ever produce paid content or "sponsored content" (meaning that the content is our idea and is not funded externally aside from whatever ad placement is in the beginning) and we do not ever charge manufacturers for coverage.
Follow us in these locations for more gaming and hardware updates:
t: / gamersnexus
f: / gamersnexus
w: www.gamersnexus.net/
Hosts: Steve Burke & Patrick Stone
Video: Andrew Coleman, Keegan Gallick - Hry
For context, here's how the PSU is supposed to work: czcams.com/video/2SleaZ68ZO0/video.html
Support GN in its testing endeavors by getting something useful in return! You can grab a GN Wireframe Desk-Sized Mouse Mat, a Red & Black GPU component mouse pad, PC building Modmat work surfaces, and more: store.gamersnexus.net/ OR you can get behind-the-scenes videos showing even more of our testing processes, many hosted by Patrick Stone: www.patreon.com/gamersnexus
"GN, you should do this testing with a safe PSU to prove it!" Already did. Check this one out: czcams.com/video/r7hNmuizMB8/video.html
Watch Part 1 (Gigabyte PSU Fires): czcams.com/video/aACtT_rzToI/video.html
Part 2 (Response to Gigabyte): czcams.com/video/Xts3pvbcFos/video.html
I shall. Once I have the money. I know it's 100% worth the buy!!
Just throwing this out there but would it maybe be prudent to use a PSU from another company in the same wattage/price range go through the exact same testing just to show that a properly built PSU can handle exactly what just killed that Gigabyte one.
@@Ravendarat They did that in the first video covering these power supplies. They covered multiple competing models that all turned off with OPP without exploding
@@faceplants2 Oh perfect, somehow I missed that. Thank You!!
TY for your consumer advocacy.
The Counter Strike intro is the best thing I have seen all week
Indeed it is!
lol at the credits
It's been a long time since I played last, but I still remember all the CS Source maps (and some early GO) like the back of my hand!
@@GamersNexus CSS best CS
Css thicc hitbox
Gigabyte be like: but the PSU was used for 600 years before you did the single cut
Hate when that happens.
I thought the same thing. GN should have started with a new, in-box PSU to prove that it hadn't been abused before testing.
@@Okusar If it works when they plug it in, it should remain working. Or, OPP should work as intended and shut it down before it explodes.
If it explodes, it's a bad unit regardless
Yeah totally unfair they used a psu that has been used once out of the box before rather than using a brand new unit from the factory.
What? Does he think the normal consumer uses the same psu for more than 2 days?
Inb4 Gigabyte demands that you use a specific PSU from a specific lot that was wiped clean before testing and was fed only the choicest AC wall power from a socket manufactured in...
You guys should make the most dangerous modern computer.
The NZXT case that caught on fire , this PSU, a 3090 playing new world (yes I know, not enough powah) and throw a crappy water loop in there to really make it jazzy .
the bad water-cooling is tactical so that the leaks put all the fires out.
@@SoloBlankets
Lmao
and to add insult to injury add a made in vietnam sticker on the bottom of the case
and the z690 formula board that caught fire
@@nowunused4631 what’s wrong with made in Vietnam
A "pre-emptive explosion" is the equivalent of "You can't fire me, I quit!"
😂 You can’t collect unemployment if you quit!
Praise the camera man ! I love how he noticed the 7 segments display flickering and quickly adjusted the camera shutter speed for our greatest pleasure. Good quality video as always !
Not that many people even know how to deal with multiplexed displays like that, much less have the camera to compensate for something like that.
I noticed that as well.
For the Emperor!
Kept on the subject while sorting it out, too!
thought I was the only one who’s ever emergency-adjusted a camera’a shutter while actually recording to avoid those annoying headlight/dashboard flickering a for videos 😂 good on him
This will now spark a competitive speedrun category and lead to the shortage of these units as everyone tries to get a wr. I am expecting the price of these to skyrocket as times get more and more competitive while the supply drops. Then desperate speedrunners will beg gigabyte "will you please, please, please make more of these?" but the hardcore community will reject anything but the original series and it will lead to leaderboard splitting.
Spark competition, I see what you did there. 😉😉
i was thinking about submitting a run to the "any% ignition" ladder but now im thinking the real competition will be in the max% category where we see how much of the unit can be melted and the timer stops when the flames go out
Waiting for fall ACDQ2021 speedruns.
I spat out my drink
I have an unopened one that I will happily donate to the cause.
It would be great to see you guys demonstrate the exact same test conditions with a seasonic or similar psu to show what should happen, that in turn would prove its not your test that’s the problem. Great work guys.
Just that with a good PSU, nothing happens. But a test where nothing happens isn't a proof. That's looking at a murderer when the murderer doesn't murder anyone. That would not prove it isn't a murderer.
@@perwestermark8920???
@@johannespeeters7368 It means just what I'm saying. With a good PSU, the PSU will shut off when it reaches the protection limit. Then it will start again after you remove the overload and cycle power. Rinse and repeat and all you see is still a working PSU.
To make it meaningful, they would need to make a *real* time laps where they continue this cycle hundreds or thousands of times just to show that it works, and stays working.
It's easy to demonstrate broken equipment just because you can capture the failure. But how to capture a non-failure? When is enough enough?
If we were talking about PSU without protection, a good or bad unit could be demonstrated by showing the amount of safety margin. But when the tested PSU have overload protection, you can't. Except to show what happens when the overload protection does *not* work. As in this specific case.
So it's bit like "how to prove you do *not* beat your wife"...
@@perwestermark8920 i think your metaphors need some work but i get that you cant keep doing it but the point is just to show what happens normally and that doesnt need to be done a lot(just like an hour) . Even if its as unimpressive as just shutting off.
@@johannespeeters7368 They have already shown Gigabyte PSU *not* failing. Because not every of the Gigabyte PSU ends up exploding.
Just that at least two have ended with something broken when it reached the overload condition and have then given an explosive failure when restarted at 60% load. I.e. the overload protection circuit itself does not cut off the load in a safe way - something breaks (not always, but often) when the overload circuit trigs.
It behaves a bit like when you cut the power to an inductor without any freewheel diode - the magnetic energy ends up inducing a large voltage that kills the transistor that disconnected the power. So the turn off looks ok - but next turn on will show the result of the zapped transistor.
Seeing him put his hand directly on the power supply while the headphone warning was on screen is the tensest I've felt watching this channel.
shout-out to the camera guy at 3:15 who probably went "i've got to fix my shutter speed ASAP" and did so flawlessly 👍
or the zero flinch from the explosion. just a couple smooth steps back.
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔.......,,,,
@@roxy_xcxc6869 i don't think gigabyte understands why they actually need "protection" with overpower protection
@@roxy_xcxc6869 the point is that if i did the same test on a seasonic PSU, it would shut down safely every single time.
@@roxy_xcxc6869 The PSU is supposed to shut itself off before damage occurs. If the PSU stays at high enough load to cause damage before hitting OPP, that's Gigabyte's fault for shipping the PSU with such configuration.
At no point do I recall GN modifying OPP; OPP was set by Gigabyte. GN was only simulating a high load at the power output.
Any % Glitchless WR PSU Death Speedrun
Actually the PSU glitched out lol
Sponsored by gigabyte
I really wanna see Steve run the All MOSFETs category now
Damn. You beat me to it. XD
12:52:17
Insightful: 60% load after successful OPP = plasma arc / soiled underwear / DOA / no RMA. Good to know. Gigabyte crossed off my build-list. Thank you!
Hanes, however, did have excellent customer service. They really got me sorted out quickly. Haven't tested Fruit of the Loom; updates to come.
@@cheeseyoger :-) ha ha.
Can you plz suggest me a reliable psu for my 1000w pc build
Can u plz suggest me a reliable psu for my 1000w build.... Plzzzz
@@BanglaUniversebd I chose EVGA Supernova 1000 G+ for my system, which provide plenty of room for expansion into water cooling, LEDs, etc. I've been running this PSU for eight months with no issues.
MY SYSTEM:
ASUS ProArt x570-Creator WiFi motherboard; AMD Ryzen 9 5000 series; PNY NVIDIA RTX A5000 graphic card; Noctua N-D15 Chroma Black cooling fan; G.SKILL Trident Z NEO DDR4 RAM. I edit video, create print art using Adobe Creative Cloud and I game, as well - Call of Duty, War Thunder, and Overwatch. Hope this helps. Good luck!
"It's almost like gigabyte uses DC Electronic load testers in it's factory", that would explain why there have been so many DOA power supplies, gigabyte have already conducted "repeatly" the OPP trip tests "for an extended period" using a "DC electronic load"
Exactly. The blatant stubbornness in their reply to GN is astounding for such a large company that previously had a good rep. They are just shooting themselves in the foot.
@@addictedtoRS They didn't have a good rep though, even going back decades. The name Gigabyte represents poor QC and failing products.
@@trevc yep can confirm, i had a gigabyte motherboard 19 years ago it failed after couple of months of use, all i did was play call of duty 1 and battlefield 1942, unsure what it was but whatever happened it fried, i still remember the smell i was only 13 at the time so i had nfi lol ever since then never touched that brand again.
@@trevc eh. they weren't perfect but they weren't exactly ECS - who is still around and making garbage. Anecdotes mean both very little (to others) and a whole ton (to yourself, if it's your anecdote) but I always held them a rung higher than e.g. Zotac or MSI (baffles me that they're popular now - they were _garbage_ when I got into this hobby) but just a hair under ASUS. "ASRock tier" or "ASUS's board was out of stock/too expensive/missing a feature" tier. And there were "better" companies that are either no longer around, or no longer making consumer, gamer, or US market hardware: e.g. DFI, EPoX, SOYO.
I think I've had more Gigabyte video cards than any other manufacturer, except for XFX in recent years. No issues with a single one of them. Bought a lot of three used ones with one DOA that I realized had clearly been used for mining after receiving it - two of them had the _wrong_ vBIOS on them, wrong DRAM type. They RMA'd it no questions asked...although I did manage to reflash the proper factory vBIOS on it first by sticking it in a machine with a working GPU. Feel _significantly_ less guilty about that now after learning how slimy they are.
My Gigabyte Z68 board is to date the only motherboard I've ever had fail on me, after ~5 years of running it overclocked, and I don't fault it entirely since the HX620 that was running it went kaboom a few months before that. After seeing this, I am now wondering if the _motherboard_ took the PSU out pre-emptively, lol. Replaced it with a beat up ASUS Z77 board and realized what I was missing. I had tons of BIOS and POST issues with the Gigabyte board from day one, overclocked or not. I dialed the same exact overclock into the replacement board and then _went even higher_ and its rock solid. Long live Ivy Bridge. Ivy Bridge is dead.
@@noxide23use 5600 with gigabyte s2h and x570 gaming X all fine
I have a suggestion to really turn the knife: make a "Certain Media Outfits" t-shirt.
I can see it making it onto the 2021 shirt.
Oh, man. I would buy the hell out of that shirt.
Or better yet, just get a consumer protection agency or a class action lawsuit to fuck gigabyte harder than they could ever imagine
Wouldn't just "The Certain Media Outfit" work?
The "N" in "certain" should be the Gamers Nexus logo.
Up next from Gigabyte: "questionable media outlet improperly tested our high quality PSU by not installing it atop the insulation pads. As explained by the verge pc guy."
Tweezers (zipties) were not used in the test and the fan was not suffocated, verge pc tester at gigabyte headquarters clearly demonstrates this is a completely incorrect usage scenario
The verge guy would probably think the power supply is a battery
@@DatGamingKid1 "some pc batteries can be expensive, but I bought this 10 million watt battery from Amazon for $5, you can choose what you use though'
Just hammer it in there.
You make them sound like the Chinese Communist Party...
This video is a titan-class power move, holy shit. Mad props to you guys for giving Gigabyte a massive slap to the face that they very much deserve.
Steve: "this video is brought to you by Corsair"
Gigabyte: "Heeeeyyyy"
Explosion speedrun. Two words I never thought I'd hear together. Lol
This is hilarious, less than a minute in, quoting from a press release, this won't end well for gigabyte.
"Explosion Sppedrun."
A surprise to be sure, but not an unwelcome one. :)
Whats's up guys, desinc here
Spoiler Alert: When there's headphones warning on the video, YOU SHOULD HEED THAT WARNING.
reminds me of mythbusters
Next Gigabyte press release: “Media outfits promoted an unintended use of our high quality power supply, which is orienting it in vertical fashion.”
The classic Apple method of "You're holding it wrong". It wouldn't surprise me at all if they did this.
"We take these problems extremely seriously. Thus, we have concluded that all of our buyers are in extreme need of education on avoiding tripping the high OPP of the PSU."
You mean like in a lot of modern PC cases? 🤔😅
They'll say that running the PSU at over 50% is bad
"plus... the members of this media outfit had their tongue in the wrong position in their mouth. That can make the PSU explode. Yeah you can make a lot of things explode. PLUS... if you hold one toe up while pushing the other toe down, you can also make the PSU explode"
I'd love to see a real-time repeat of this test with a comparable known good PSU so we can directly see what the expected behavior is. Loving this kind of in-depth content, keep up the great work :)
... Nothing would happen with a real PSU
I've seen a lot of people say this in the comment but the PSU would just... turn back on like normal. Ideally you'd be able to repeat this test multiple times without failure as OPP should be tripping before the PSU is actually damaged. They even said in the video many of the PSUs were failing when the brought them back up to OPP, not when they were turned on to normal loads.
well your request has been granted.
they did the same thing with a brand new PSU from EVGA, that came straight out of the box.
spoiler: it didn't explode... as you'd expect
What’s a real real PSU? The best thing on the market?
I had my PSU (seasonic 750 titanium) going in short circuit protection under certain loads because (after weeks of trying figuring it out blaming bios settings, LLC voltages, usb devices etc... The pc just turned off sometimes after hours, sometimes never, often while alt tabbing with many things open like discord, game, overlays etc...) a naked metal inductor of the vrm was touching one screw of the monoblock... It saved all the pc at least 10 times before I could figure it out the real problem.
Then there are PSU like this.
I wanted to say when seeing this video: Seasonic is the Bentley of Power Supplies! I'm unwilling to use anything else.
A part of me hopes Gigabyte keeps denying the need for a general recall, so that GN will keep making savage PSU video's.
And so that a more formal recall may be in order. I love when companies are forced into doing the right thing, rather than being able to deny, deny, deny and then let up at the last second and call themselves the good guy.
I have been watching videos recently, by two guys who have lived there for more than a decade, about China. Culturally Taiwan is part of China. The problem is, they just psychologically can't assign blame or take responsibility, unless forced. This is the same with Thermaltake's bad case designs. Nobody is allowed to say that there is a problem. The guy who designed this thing, and possibly the guy who hired him, is already gone.
If gigabyte keeps denying GN should take a bunch of the same PSUs and make them race each other to death.
Gigabyte: "I reject your reality and substitute my own".
Steve: "Steve wants big boom".
Gigabyte: "Am I missing an eyebrow?".
@@krazycharlie smh
the counter strike SFX really enhanced my experience.
It’s ATX
(it's a joke)
@@MelterNeXus ?
@@immxturelol sfx can also refer to a smaller psu, as they generally (for computers at least) come in two sizes. SFX and ATX.
@@oleanderthor-borre9506 Some SFF PCs have TFX or Flex-ATX PSUs.
@@nathangamble125 ahh, well now I know. Thanks 👍
Safety glasses are highly recommended in working with exploding stuff. Otherwise, damn good job guys. Thanks for everything.
The way Steve smiles at the beginning, his eyes really sparkled like the P750 when it exploded.
I love that this thumbnail FEELS clickbaity but is, in fact, 100% representative of just how bad this issue is. 😂
This is why GN is the best testing channel, if there's an issue, they put in a fuckton of effort to prove it and show that the companies are truly BSing us.
(also hi badseed love your videos)
It's reverse clickbait. It baits you in and completely delivers.
Oh man, you're a fan of GN too? BadSeed + TechJesus collab when? xD
sometimes things are that representative :D
I xp'ed that some years ago with a 500w psu. Loud bang and it tripped the circuit breaker. This was night time, so sitting in complete blackness and smelling burnt electronics was not fun.
GN: "This video will be uncut."
Me: **checks video length again** "oh this gonna be good"
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔 ...........lol
@@roxy_xcxc6869 lol
@@roxy_xcxc6869 I'm not an electrician but I believe that no, they would not explode if they're well designed, but good try Gigabyte employee
@@roxy_xcxc6869 lmao did you even watch the video? A well designed PSU should see "5000w" and safely trip OPP and shutdown. Once you turned the PSU back on it should work as normal and not explode.
@@roxy_xcxc6869 lmao you bodoh or what? Don't understand the purpose of the video?
Terrific video! Thanks for the testing. I own one of these that is within the serial number range they posted, and I want to sincerely thank you for the head's up on the potential failure probability. It prompted me to replace it before a failure, and might just have saved my components. Both Newegg and Gigabyte have been responsive to my inquiries into the "return and replace" mentioned in the Gigabyte press release. Hopefully I will get a correctly working PSU shortly.
What I love about this is that you are literally experts in this field, performing real world experiments, and the manufacturer is adressing you as if you are just some guy with cameras.
We live in a world were data, facts, and research dont matter. All that matters is who you choose to beleive.
I can't wait for Disappointment build 2021, I have a feeling I know exactly what riser and PSU are going to be used.
Fireworks
The only question is which would take that build out first, the fire or the explosion? 🤣
Depends, do you put the emphasis on 'disappointment' or 'build'? I feel like combining a GB PSU with an NZXT riser in one system would rather result in a disappointment demolition rather than a build...
Aha but the gigabyte PSU will fix the NZXT ryzer cable, cant catch fire if the PSU bites down on the hidden tooth cyanide pill first
*deflagration build
Gigabyte: Noone is petty enough to actually prove us wrong
Steve: It's not about being petty, _it's about the truth._
It’s like Linus said. Steve is a dog with a bone.
Gigabyte; you can’t handle the truth!
@@VladMcCain heck, I wanted to comment that hahahaha.
And also about being petty.
Gigabyte is petty enough to issue DC load testing is "not real" and refusing to honor RMAs.
An eye for an eye.
After finishing electrical school videos like this are super interesting to me. Thank god there are guys like you that are willing to take the time and risk the burns to test these things to failure for the rest of us
Just got back into PC building since 15' and so glad I picked Evga for my PSU & GPU. Had not seen your channel at the time. Was tempted to buy a Gigabyte MB since the price was lower than comparative ones but Gigabyte was the only MB I had had issues with in the past so went with Asrock instead. I now understand why the price was lower. Thanks for all your hard work and you have a subscriber for life now. Now off to buy one of your cool pads. Thanks!
GN is a tour de force in this industry. NOTHING gets past these guys. Impressive work and invaluable to us as consumers.
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔.........xD
@@roxy_xcxc6869 You should start paying attention more, it clearly explodes only when under smaller loads and does so numerous times. If the 1050W was what was killing it, how come none of them exploded at that stress point?
@@roxy_xcxc6869 they've saved us all so much money and time thank you GN!!!
@@roxy_xcxc6869 The fact that the psu literally explodes should leave gigabyte responseless and immediately call every psu out there. A psu should never explode no matter the wattage.
@@roxy_xcxc6869 wait what are you trying to say? It was failing under lower wattage
Gigabyte: "I don't want to name/fight you!"
Steve: "I wouldn't wanna fight me neither."
A Raimi meme in the wild
@@EpicGamingEct but then it "trips" below what is advertised max power after hitting opp once
Thems fighten wordz pardner!
@@EpicGamingEct Bro, you having fun copy/pasting this comment into every top comment thread?
The problem isn't that it tripped at 130%, the problem is that it exploded at 60%.
"Your already dead."
Great and professional, as always. If I've been looking for a new PSU, this is exactly what I would like to know before I make a purchase (in this case a risky one).
Thanks for sharing guys!
When he put PSU sideways I was thinking of those mines with label "towards your enemy" for direction of the blast :D
🤭
claymore
I love how you guys will continuously hammer them harder when they don't admit to their mistakes. This goes for any company you guys have reviewed and will review in the future.
Yea, Gamers Nexus is a blessing to the entirety of the tech community. I really wish they were even more well known so more people could be better informed about these important things.
"Just repent and we'll stop lashing you."
Holding Gigabyte accountable is the only way to make them correct this. A pat on the shoulder isn't gonna roll here.
Pretty sure Steve's just mad that gigabyte refused to give them anymore freebies. I have the same exact PSU going strong for over 2 years now.
Imagine how few people would know about this if Gigabyte just immediately offered warranty returns and didn't dispute the results. There'd just be buyers complaining in reviews and one video saying Gigabyte was fixing it. So dumb.
I knew that boom was coming with all the headphone warning signs but it still startled me
I was shocked at how load it was too.
Yooo dapz
I was actually in time to turn down the volume. But then the message disappeared and I thought I'd missed it so I turned up the volume again. I think I need new pants too...
They should've added a timer instead of having the message disappear thinking we're in the clear.
@@triadwarfare indeed. Countdown from like 15s.
Thank you for the good journalism and test work!
Finished my first PC build back in March, uses Seasonic for the PSU (after a bit of research on the internet)
And this is why people should do some research rather than choosing 'brand'.
Thank you for making the video, great job as always
Might be a dumb question but if that testing machine is drawing power from the psu what does it do with all those Watts? does it just heat up?
Not dumb at all! Yes, it basically turns into heat. You can hear the fan ramping really hard at some points during testing because the test machine is attempting to shed the heat. Very good question!
@@GamersNexus Back to you, Steve!
@@tiestofalljays Thanks, Steve
@@GamersNexus heat: aka basically the only electric energy conversion that is literally 100% efficient!
@@GamersNexus thanks for the answer! 👍🏾
I love when this happens (companies trying to discredit Steve and his team's findings), because Steve goes full savage and shows how dedicated he and his team are to the community. They look badly behaving multi million/billion dollar companies straight in the eyes, kick them in the balls, and then absolutely destroy them when they keep up with their b.s.
the entire industry fears this man
What's even greater is he clearly shows they aren't just putting massive load on one connector.
He's showing they opened up some of these PSUs to figure out the rails and created loads for every single rail in the PSU separately, by connecting all the necessary cables to all rails.
That's dedicated and perfect methodology.
There are a couple of problems with Steve's testing.
(A) Firstly he loaded this PSU for a prolonged time above its rated power. You might think that 2-5 minutes is short time but it actually is very long time in relative terms. These power supplies are not designed to expect a sustained load above their rated value but instead they are only expecting brief spikes. Spikes can be created because of inductors, etc and they only last a few ms. So although during their service life these PSUs may see the power shooting momentarily to 1000W or 1200W they are definitely not designed to see the load remaining there for seconds let alone minutes. That is hundreds of thousands of times longer time than expected.
(B) More importantly when the failure occurred he essentially had a big step-function current (or equivalently power) draw. The load went from 0W to 450W in a microsecond. Such high current slew rates (rate of increase of current) are incredibly damaging to the components of the PSU. In real PCs, both the motherboards and the add-in parts (GPUs, etc) are designed in a way that limit the current slew rate to exactly avoid such step functions which can lead to a catastrophic failure. What Steve and Patrick should have done was to ramp up the current/power in smaller increments or use an inrush current limiter (which can simply be a thermistor) between the load and the power supply.
So yeah Gigabyte is absolutely right on this one. Not saying that there aren't other design issues on Gigabyte's PSUs that can lead to catastrophic failure but Gamers Nexus' methodology (especially the huge current slew rate) is just way too brutal. ATX PC power supplies are simply not designed or validated for such a DC Load Step Response.
I'm waiting for the Blue Origin style infographic showing that his shop doesn't really exist and his stuff is unsafe.
1:23 They just took note from the Portal 2 ending: "Reactor explosion timer destroyed. Reactor explosion uncertainty emergency preemption protocol initiated: this facility will self-destruct in two minutes."
Gigabyte’s PR is the same as any catastrophe in a dictatorship: deny, deny, blame someone else.
Corsair really missed the mark with advertising their PC case, this was literally the best opportunity for their PSUs 😂
I doubt Steve would agree to that.
It could make him look biased.
Last year my VS550 exploded with better sound than gigabyte 😐
@@gauravmohore3034 oh, oh no, thats the same psu i have.... should i be worried? i had it go off on thermal overload once or twice but i put that down to the fact its pared with an fx8350 and is in a shit case
@@firenado4295 check the PSU serial number, gigabyte will swap it out if you're in specific serial number range. also make sure you have this exact model, not all Gigabyte PSU's have this issue.
Nah my ax850i can't power my system correctly opp kicks in at a small load.
I can't imagine what crosses the minds of these companies when they see a GN expose and are just like "Double down! He'll back off!"
Out of touch PR people. Speaking as someone who has worked in fortune 500 companies, the people who decide on the response are generally in their late 40s and 50s and are completely out of touch with the new age mentality. Most old timers in these companies think they can lie their way out of incompetence and the "system" will protect them...because most of the time it does.
"In real-world, no one is going to use their PSU in the way that they have tested. Therefore, all the Newegg and Amazon reviews saying this PSU blows up are false. We will process the returns."
Also Gigabyte: "Serial numbers don't help cus I can't read."
Source: Reddit user saying that their Gigabyte PSU was not eligible for refund... for some reason... despite having the same serial number as posted with the refund terms.
@@mistere5857 Good news about the internet is that people are getting more and more informed. It's easier to just lie harder, but there has to be point that "lying is not lying if you don't get caught" certainly is getting harder and harder to the point that being a decent company is much easier anyways.
And much more profitable to maintain their reputation.
@@chrinaldi Noctua is an example of a company with good reputation. Even tho previously a shitty colored one.
@@chrinaldi in china all your refunds are belong to us
As soon as saw this video i changed the power supply p750 immediately, thank you man.
Keep up the good work.
Gigabyte uses small, closely calculated mosfets that have high resistance causing it to heat up alot, they could change their mosfets to bigger and low resistance ones or just the heatsink and pad and change their ocp because this on is bad af but ofc they wont do anything.
I pulled this same PSU from my daughter's gaming PC this weekend. The Gigabyte PSU was part of the NewEgg shuffle for an RTX 3070. I'm hoping GN's investigation will lead to a recall and a refund.
Doubt. Gigabyte is way too cheap and Chinese HQd to get anything done. Not in a discriminatory way, following in MSIs footsteps (although they are Taiwanese) by never admitting negligence. And their US offices can't do shit right now.
I got one with my gpu from the shuffle as well. Glad I never opened it up. ^^
Same here. Except I saw the issue videos before the psu was deployed. It’s still in the original plastic wrap on a shelf. Not sure if it works or not. Or even what the serial number is. Worst case I make it a bench top psu for hobby electronic projects where I don’t care if I blows up.
Wise choice
Newegg burn your house down shuffle
Some say that the fireworks that lit up the sky on the evening of the Fourth of July were actually a series of Gigabyte PSUs exploding simultaneously.
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔......xD
@@roxy_xcxc6869 OPP isn’t user configurable, this is how Gigabyte spec’d these power supplies. You clearly have very little clue as to what you’re talking about so please just keep your mouth closed until you take advanced electronic design and math classes before making an ignorant statement
:D those nut guy.., use 750w psu & run at 1050w omg crazy guys. Keep bleeding the transistor inside again & again until make them explode on purpose lolol 😂 you try use msi psu then & bleed it with 5000w opp test a few time & see msi psu will explode or not..hahah try it....,
this is how Die Hard 4 actually started...
Oh lawd they won't live this down either ....all they had to do was to admit they fucked up
"You down with OPP?"
"Yeah, you know me!" -Gigabyte, probably
I hope the brilliance of this comment doesn’t go unnoticed.
It's noticed... excavated, dusted and still judged funny after all these years 👍😊
Well done :)
My wife bought a gigabyte P750GM a year ago for a system she built.
This weekend it blew up. We thought maybe dust bunny shorted out something so we ran some air through it visually inspected, didn't see any blown caps or magic blue smoke burn marks.
I found this video today after I tried to test her PSU with a PS tester (no load just voltage check) and hitting the switch resulted in an immediate arc in the PSU.
Put her SSD in an old system of mine so we know they survived, but don't know yet about her GPU, CPU, MB, or RAM. I have a known good PSU (EVGA 850) to test with. Fingers crossed it didn't nuke anything else.
So nerve-racking watching this video. Every time Stone put his hand near that power supply I was like, "Noooooooooooo!"
I about had a heart attack when I saw him holding an "exploded" PSU in his hand waving it around, until I realized that was the "fresh" one that is next up for the executioner's block.
And then I stroked out when he went back and ripped open the exploded one in the same take. He's for sure WAAAY smarter than me about this stuff so I'm sure it was really fine, but I didn't see/know whatever happens to super make sure capacitors are grounded out. .... or maybe it was just a "yolo, don't touch the deranged innards" and trusting yourself not to fuck up.
Not the left hand please, please have a habit of touching electronics with the right hand. (Electrocuted pathway)
Especially many times he covered the fan with his hand.
... professional safety gear: rubber flipflops for insulation 🤣
This felt like watching a horror movie. Where the blonde is hearing strange sounds in the other room and tries to open the door and you are like "STOP...NOOOO".
Gigabyte PSU's are like having the spirit of ElectroBOOM in your PC. Also, Linus should send Patrick some free underwear for exhaustive testing.
i don't know if LTT Store has brown underwear or shorts
Stopped watching electroboom because he had the ad ho and scam enabler on there.
lol
@@m4c1990 He had who on?
Dude, thank you for this video! you stopped me from buying a Gigabyte PSU, now I get why they are priced to be such a good deal!
Your Technological Journalism is unparalleled. Well done
“Can I get you to touch this capacitor?” Should be a GamerNexus shirt.
I second this.
I need you to discharge this. Across your body.
When I was doing a traineeship for electrotech, my 'trainer' tried to get me to discharge a new wrapped fridge, but I could feel it jumping before touching the frame.
Then he came along and got the zap.
Sucker.
Or: "It's not even over 9000!"
So, because you use an electronic load, would this be considered a Tool-Assisted Speedrun(TAS)?
Nope, as they need a load it is seen as necessary equipment. As you need some sort of controller/mouse/keyboard to speedrun a game.
I mean, sure, since Gigabyte sure seems to be acting like a whole bag of em!
It would've been TAS if they pre-programmed the sequence of actions into their testing device and executed it with a single "start" button.
@Edge64 Damn son
"Steve Burke would hang on to the record for the remainder of 2021, until a old challenger joined: enter Matt Turk." -Summoning Salt
The security cam video made me realize how calm and committed your camera operator is. That's awesome! You two flinched but they just kept filming.
You can see the legitimate fear in the security footage when the PSU blows up. It's so damn unprofessional and childish for the company to respond like this when people could get seriously injured or even die from something like this
This sort of nonsense seems to be more commonplace in the days of "Global shortages". I bet they are using lower grade Chinese parts and passing them on to consumers.
Or don't push a PSU wayyy past what it's rated for
@@Kadori328 that’s what the opp is for tho, works on literally 90% if not more psu’s 🤣
@@Theredstar2478 still don't put your faith in it as your guardian angel I'd you catch my drift.
@@Kadori328 well yea for sure but I mean on other psu’s I’m guessing it’s been actually tested
"Gigabyte was afraid to name us. It's OK, I would be too if I were Gigabyte." Damn Steve, gore warning first!
its a corporate response.. so lawyers do not have to get involved - its not about fear.
@@ProcessedDigitally yes people think these massive companies can simply say whatever they want.
@@ProcessedDigitally So out of fear? Got it.
@@ProcessedDigitally Well being purposefully vague in responses that aren't audited by lawyers is OUT OF FEAR of being sued. So it's standard practice OUT OF FEAR!
@@Celiktaban Have you ever heard of an airline called Ryanair? When customers complain about them on Twitter the CEO just replies with: "We're the cheapest for a reason. No refunds!"
You guys should make a ‘most dangerous pc’ with a gigabyte psu, NZXT H1, and a bunch of super underwhelming components
11900K on a throttling garbage Z590 board
Edit: how could i forget, Dell’s 1660 Super for the GPU
@@depth386 11900K with stock Dell prebuilt cooler.
EVGA RTX 3090 playing New World
@@EricdaRED95 You sure you wouldn’t do the 3080 Tie instead? I mean sure you can knock the “value” of the flagship all you want but the MSRP of the 3080 Tie vs the 3080 was a real sour worm
They gonna get knocked on the door from FBI...
I just replaced my faulty, old PSU. It was triggering some protection (Current, power, thermal idk) every time I loaded a heavy gaming scene, making my entire PC shutdown instantly and restart. Unsure what the problem was, I thought it was my mixed RAM at first, I did a lot of testing and used my PC for a week or 2/3 in this state for work and lighter games. It crashed like two dozen times before I replaced the PSU to find that that was the fix.
Now just imagine I had a PSU with a fault like this. I could've blown my entire PC. It's just insulting to see Gigabyte denying there is a serious problem here.
And btw I almost actually bought a GP-P750GM as replacement, but didn't because it had low user reviews. So thanks reviewers.
3:27 is really cool - you can see the frequency of the psu tester (and its display) go from 60hz, with significant banding to a smooth 50hz, also what the camera uses for anti banding
How bad do you want the PR-disaster to be?
Gigabyte: YES!
Oh man.. I hope they put more effort into their mainboards..
I used to have Gigabyte motherboard and it was all right. Currently I use their 27' monitor since 2020 and I am very satisfied.
@CALLER ID actually I never updated the firmware because it is hooked to my XBOX. How do you even update it, just curious? You download update file from PC and install it?
@CALLER ID thanks for reply. I see. Thats so unprofessional of them.
Yeah, not feeling so hot about my Aorus x570 anymore.
@CALLER ID Yeah, ASUS also have their moments
BREAKING NEWS!
EVGA will be renaming one of their PSU lines - they say Gigabyte can now use "Supernova" for their PSU's.
Right there with the "Fireball" HDDs and "Firestorm" cooler.
MSI came first and removed the whole "lightning" series :D
Terrence & Philip from "South Park" are giving up "Assholes on Fire"
LOL.
EVGA should name their PSU White Dwarf, it'll be putting out energy steadily for billions of years.
first I felt bad for Gigabyte, then I realized that I have this exact model in my build.
oof
Me too
Time to get a different PSU lol.
It's a feature Steve and it's called First Intervention Reset Explosion. Also known as FIRE
The silver lining is you could use this time to introduce GN branded safety glasses with the word Gigabyte as a discount code.
I know it probably won't make too much of a difference but can we get Patrick some safety goggles when he's working on something he's fairly certain will explode? lol
...positively no flipflops 🤣
why?
Flipflops & electrical testing shouldn’t be done together. Then again being able to do an explosion speedrun on a power supply isn’t something that should be possible either. But look at gigabyte accidentally creating a whole new competition for youtubers.
More fire extinguishers on set as well.
Better an anti-bomb suit if he need to disassembly those psu
Just the intro alone had me in stitches!!!! Nice one Steve.
Oh Gigabyte is going to nail you for using a variac, they will claim it supplies an in consistant variable voltage that the PSU is not made for and that is the cause for failure they won't say anything about it galvanicly isolating the PSU making it safer..... Watch for it to hit the next press release, they will say anything they can to sidestep this clear bugger up that they made... Just an FYI, A DC electronic load can be a PC a laptop or a DC motor....... A load is a load, wether it be a PC or a damn resistor...... It's a frikken load..... Just own it Gigabyte you messed up.... OH! and testing to OPP boxing it up and giving it to a customer having it DOA.... Nice Gigabyte....
Any update on the testing for the "good" serial numbers for this psu? I have one and am waiting to see what the results are for those. Love your content! Keep up the great work!
I can already hear gigabyte's response: "It's not brand new in the box"
Omg how could they use PSU longer than a few days they were designed to live, who cares if it kills all your components after.
And ibuypower came out of nowhere and defends gigabyte and everyone already knows ibuypower pre builts are horrible value and dangerous
@@EpicGamingEct your english and your sorry spam of a comment is not credible lol
@@EpicGamingEct Clearly you didn’t see the part where the PSU literally blew up
says the company that DOESN'T SEAL THEIR BOXES
If "certain media outlets" doesn't become a t-shirt print GN is dead to me, 😂😂😂
New line of under garment wears.
It's "outfits" which makes it even better lol
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔........xD
It'll be the line for Aug- "certain Media outlets" for the Year End Biggest let downs shirt.
Dude that lens change looks great!
The "Patrick Stone: speedrunner" tag gave me a solid chuckle
: D this guys is nut. Running 750w psu at more than 1050w! Of cause will explode lah omg ? Which means we can set opp at 5000w & see then. See whether msi psu will explode on 5000w opp then 😁🤔🤔.........xD
@@roxy_xcxc6869 Won't explode, because the OPP will trigger and safely shut off the PSU. You would know if you weren't bo lan tiao. Gigabyte sai sia.
@@roxy_xcxc6869 you've spam commented this everywhere. You're very ignorant, they're testing the safety mechanism not the power load.
"The explosion when approaching over-power protection is to prevent the power supply from exploding from over-powering." - gigabyte, probably
big brain move right there
I bought one of these damn things 2 weeks ago. Glad a ran across these vids when I did. Swapped it out today 👍🏻
Extremely informative. Thank you.
Years ago, I worked in a factory that built large back up power supplies. The test phase of the bigger ones were run at 100% load for eight hours then, iirc, 125% for 4. The entire factory would be in the low 100s (degrees Fahrenheit) that morning. For some clarity, they would run these things starting on a Friday and ending Monday morning around 4am. The worst part was when one would fail a Hi-pot test. That little flare from that 750w power supply is nothing compared to liquified 1/2” bus bars and 3-6” plasma flares. The smell of the other fried components was atrocious.
I guess the point of all that was the testing they did wasn’t considered complete unless the units could put up with the worst the end user could throw at them and then some.
Clearly, this manufacturer has a design/component/quality control/budget issue. It would seem that their language in their reaction would indicate that they know it.
IF they perform any QC to begin with..if
Kind of reminds me of another "Gigabyte-explosion": I used a beQuiet! power supply on my Gigabyte G1 build couple years ago (Z97 G1 + 980ti G1). The caps on the mainboard exploded and the G1 burned to a crisp - will never forget that smell. Once the cooling was removed, I gently blew some air over the nvidia board: some burnt ICs simply dissolved into dust. They just disappeared from the board. The power supply is still kickin' though, as is the mainboard. Replaced the fried crisp 980ti with an RTX OC on second PCIe. That smell though...my my...
@@matthausadamitz2127 Yeah, the stuff you from that can stain your sinuses for about a week. It’s awesome, especially the 3rd time.
This testing to 100% load applies to a lot of things. Even some engineers don't know that critical components can be and in some cases should be tested to 100% load befpre being put to use. Not to just test as a sample. This could be mechanical parts like suspension parts of a racecar, cable ties, welds or joins between different materials for example.
Some people also think testing is somehow always destructive. It isn't. Testing at 100% load is not the load at which the thing should break anyways. It is the maximum which the part should handle without issues. More to the point, everything is designed with a safety margin. Elevator cables for example have a safety factor of 11. That means it needs to be able to handle 11x the maximum design load. Most of the time the safety factor is 2 or 3 for most parts. However things should always work at 100% load and not break. Be that 100% load a weight, amperage, watts, torque, heat or speed.
@@erwinlommer197 Yeppers.
I'm sure its been suggested plenty at this point, but I'd love see Electro Boom get in on this, this seems right up his alley.
These power supplies are even too dangerous for Electro Boom.
yeah, I'm pretty sure you can't safely shock yourself with those
Was going to suggest photinicinduction but gigabyte already did a fine job making sure that those PSUs will fail with flames. 😂 On a serious note, sending one unit to Dave would be great. Maybe he wants to dig into it more.
Hahaha
He will probably put his junk into one of these... "Let's see if it's really explosive!"
This is why I support Steve. Tech companies try to bully reviewers and media outfits but Steve calls out BS and defends his reputation.
You know you're badass when you let the Gigabyte PSU face you on T-shirt, shorts, slippers and no eye protection while clicking the wattage up up and away.
As a computer engineer, I don't have those balls of steel.
I have a feeling the Gigabyte Christmas Card this year is gonna be one of those 'this message will self-destruct in 5 seconds' types.
Ah, so they'll just send more power supplies.
Not sure they really want Steve blowing up more of their PSU's on camera....
Just without telling you so 🤔
Patrick putting his hand on and near the PSU each time is edge of your seat nail biting stuff
Meh even if the case shorted to live, 120v makes you recoil, not grip on... I'd be much more scared of those PSUs that DON'T blow up, sitting there with fully charged capacitors...
@@HMan2828 Indeed, the fun is what I did back with an AT PSU. I picked it up while it was powered up, not realising there was a cut-out underneath without a plastic shield, so I touched the PCB likely on the AC side.
My left arm was numb for an hour. I was just lucky it grounded in the same arm rather than going across my chest. Still amazing that despite being a short across my hand, I felt it up my entire arm.
every PSU makers: we try our best to exceed 80plus standards.
Gigabyte: we have exceeded 80Plus Uranium standard.
If you had just a resistive load, or just an inductive load, or just a capacitive load, each is different than the others. A DC electronic load placed on the PSU should be a combination of all of them. Old load benches (which I used to use probably before you guys were born) were strictly resistive and a LOT different than the newer and more accurate electronic loads. Certainly, if the PSU is properly designed, it should OPP trip at a certain point, and do that every time for many more times than the 2 shown in this video. (NOTE: I ran a PSU lab at NCR years ago and it was my job to perform PSU qualification testing and advise the engineers what failed and possible fixes. Since then I've had a LOT more experience with electronics and computers.) In short, this is a horrible failure and GB should have NO excuses and if I or other engineers had let something like that go at NCR, people would have been fired immediately.
Honestly, this sad for a company to do that to their customers like thinking about this from another side.. if we didn't have any tech channels reviewing products, more people would have lost their stuff and our complaints would never be heard. Thank you steve and keep your amazing work up
Yeah totally
Not to mention potential house fires and deaths because of such negligence in this product.
@@Leignheart True, My friend actually lost a motherboard and GPU cause of it, I hope people by now are aware of this psu for safty.
@@meownime1603 Especially when they're so many people that underestimate the importance of buying a good PSU so you won't have to deal with that shit, i can't count the number of times i've seen recommandations on hardware where people just put everything on GPU but don't seems to give a shit about others componants and PSU is the part you don't want to cheap out.
I would have fell in the floor laughing, if Steve said the imminent explosion of this PSU is brought to you by Corsair, lol.
I mean corsair must be happy to see GB fall flat on their face.
I snort laughed when i read this so you win
@@NoxIF34 Gigabyte and Corsair don't compete on any products from what I can think of
@@andy_byrd PSU's, Prebuilt computers, keyboards, mouses, headsets, SSD's, Memory kits. Gigabyte getting bad press is good for everyone else in the market.
@@andy_byrd case's as well
Lighting looks great in this video Steve.
In a modern MOSFET configuration like DrMOS on mainboards inside the IC package is a build-in sensor that protects the device in case of sudden overcurrent and overheating. But on conventional MOSFET such protection to avoid explosions is not available. The design on the PWM controller is crucial because the detection of sudden overcurrent on the primary switching MOSFETs is the most important.
In a nutshell when both mosfets are conducting at the same time it's a short circuit. One mosfet must turn off so the other can turn on. If both are on at the same time kaboom. A well designed power supply also reads the mosfets temps and if they get too hot the power supply shuts down anyway regardless the OPP. Close to 200ºC is too high
Seems interns have been put in charge of their factories.
@@damionb7946 smells like you don't know what you are talking about
there's many things that could kill mosfets and one of them is cross conduction from having not enough dead time as you said. if the problem was dead time then paralleling mosfets won't help.
@@punktkomma9489 How do you conclude that? It absolutely is possible to run mosfets in parallel. If you want them to run absolutely perfectly in parallel, you'd need to miller plateau match them all. Most applications can handle a few microseconds switch on time difference though.
@@n.shiina8798 Just better mosfet gate driver and bleed resistor, or better mosfets.... or both
I love Gamers Nexus. Steve seems a real guy who loves the kitty cats and looks out for the consumer - short of the cape he's a hero to me.
Not all Heroes wear Capes
Give it time, the hair will become the cape.
Thanks Steve.
Not to sound pessimistic but always leave room for doubt/skepticism, even if they are heroes.
@@mab2187 calm down.
Even tho this is such an old video and I know Patrick is fine, just watching him with that PSU right next to his arm sticking over the edge of that desk continuing to touch it and shove it around, pointing it towards him in T-Shirt and shorts really had me hissing:"Stop touching it!" in concern by the end lol
Can we stop for a minute and appreciate how cold-blooded camera guy is
I own one of these P750GM PSU's, used it with a 3080/5900X for about a month, PC would randomly crash so I swapped it for a different 850W PSU and never even thought twice about it until I saw this video, now I'm getting emails from newegg to return for a full refund, thank you GN! ❤
You should tweet / send a copy of that email with your PSU serial etc. to GN's email - I am *sure* they will be interested
Same issue for me random crashes I sent mine back at xmas already tho glad it didnt explode all the same
I love that the reaction to Gigabyte talking sh*t is basically "f**k it, we'll do it live".
*uncut
Simple answer: I will never buy GB PSU. Ever.
I love how the CSGO intro is actually putting in the accurate code
I love how the credits are all callouts, the last one is literally me every match