nvidia is smart tho, same thing with 4090s, they push 600 watts through a 16 pin cable while it should be a 24 pin one, they melt, and the customer needs to buy another one, amazing
Replace your barrel jacks with XT30 connectors. Its what I'm using on my laptop now and it is 1000x better than the stupid barrel jacks that fail so often. The XT connectors are polarized so there should be no way to connect them reversed.
@@bencromwell4453 60amps might be overdoing it. XT30 is 30 amps but probably better to downscale that to 25 amps. At 12v its more than enough for testing.
@@bencromwell4453 There is no reason to use XT60s. It's a waste of money and a bulkier connector that will be absurd-looking with the wires running to it.
@@jorgealzate4124 yeah, I don't think he'll be pushing over 15amps on that test unit with those wires anyway. The XT30 connectors are capable of 30amps continuous if you can fit 10awg wire on them. The wires heat up more than the connector does and even fitting 14awg wire on an XT30 to get 15 amps continuous is difficult. But I don't think he'd need that much current on the test rig. Either way, XT30 or XT60 connectors are the way.
The 150W from the spec on the 8-pin PCIe power connector is quite conservative, that is part of the reason why they don't usually melt even when not perfectly connected. But remember one of the 8 pins is used for presence detection instead of a positive, so only 3 (x2) pins are actually carrying current. When all the 8 pins are used there is a bit more headroom. Electrically, I would say they are still ultra-safe for about 225W, so this is reasonable for the TDP. 12VHWPWR is actually the opposite, it "walks on the edge", and a small manufacturing issue or improper seating can actually do damage. Great video as always.
What does PCIe pull, like another 35W? If the max draw of the card really is 300W that's still not enough, maybe for transients. Only Nvidia would design a $5,000 card but allow some critical design flaw to be given a green light 🤦♀ Just looked at the Furmark test at the end, card is pulling 297-299W. Assuming the most robust 8-pin possible, 3 * 8A * 12V = 288W Yep, not enough. Nvidia be goofin 🤷♂
@@quantum5661 Yea according to PNY blog that's what the adapter is for, dual PCIe 8-pin to EPS-12V 8-Pin which is rated for 300W. It's still kinda nuts to me they wouldn't just make the shroud & PCB slightly larger and use 2x 8-pin for a $5,000 card.... Maybe there's some server spec that required the cards to be under a certain size, I have no clue, but they're cutting it close either way. I always get a PSU that's a good 100W over what I need, because its safer and keeps the PSU cooler. Pushing things to their limit tends to invite things to heat up and burn lol Safety margins = 👍
@@The_Man_In_Red pcie connector can provide 75W according to the spec, and it indeed does since usually a couple of the phases are powered by the slot to even the load. In the case you describe 299W < 225 + 75W. As I said, opposite to the case with 12VHPWR there is a hefty safety margin in all thode numbers. Transients can be twice as that. Nvidia does many wrong things, I agree, both in design and price. I just do not think the connector of this particular GPU is one of those.
@@Igbf just to be clear about the 75w pcie slot spec: you only actually get that number if you combine all the voltage rails, by default its only 66w sustained.
Based on your experience, I also found out that extreme heat liquidates the cables and allows them to be bent into any shape or form you can imagine, makes working into desired paths very easy
They're powering it by using an EPS connector, not a PEG connector. Notice the four 12v leads rather than three. Then they're using a dual PEG 8-pin to EPS adapter, using two 12v from each. The EPS connection is "rated" to 300 watts, though I'm not sure where that rating comes from since Molex Mini-Fit Jr. 8-pin dual row is rated at 10a per 12v terminal on 16ga wire for a total of 480 watts, and on 18ga it's rated for 8.5a for a total of 408 watts. Since it would be the same for the PEG connections, you still have four 12v terminals, giving you 408 watts... worst case scenario. So, overloading the connector wasn't the reason for the failure. Also, I would like to point out that it's a common misconception that the 8-pin PEG connector is rated for 150 watts where the 6-pin is rated for 75 watts. Those are PCI Sig sensing specifications and they have nothing to do with how much current they can handle. Mini-Fit Jr. predates PCI Sig and even GPU's needing supplemental power. Hell, the first GPU's that did used the "oh so wonderful" Molex 4-pin. The 8-pin and 6-pin are actually rated the same, so with 16ga wire you're looking at 360 watts for either or 306 watts with 18ga. Even if they had used a PEG connection, they wouldn't have overloaded it, so long as they ignored the sensing specifications and use 16ga wire. With 18ga, it would be pushing it... kinda' like 12VHPWR these days.
@@PaulBlartGaming He just doing easy jobs and if you send your device and he decides to not repair it just because then you just pay alot for nothing. That's he case.
I would have spent the wires before putting them through the solder holes so they would be at a 90° angle and you wouldn't have had to put that much torque on the solder joints. I saw someone doing that with a circuit board and I thought it was a great idea and I kind of thought you were going to do it. All this stuff really interests me and I'm so glad I get to watch you, man.
What I learned from the video is measure twice and cut once... well relearned that lesson? I guess I learned that a rule of thumb for construction applies to GPU repair too. But then again it is applicable for anything you need to cut to a length.
Hey there Tony, the connector on the card is actually an EPS 12V connector, which is if I recall correctly rated for 300W. (Not 100% sure about the rating) Anyway, because the connector isn't a PCIE 8 pin connector, NVIDIA includes the 2xPCIE 8 pin to EPS12V adapter.
Side Note: A lot of 8 pin PCIE cables from power supplies have quite a big safety margin in them, so they are oftentimes able to supply more than the rated 150W without problems
And the PCI express slot can deliver up to 75 watts as well. I used to have an Nvidia 1050 GPU that would get all of it's power from the slot, no cables required.
The connector is EPS12V (300W, 4x ground 4x 12V), the same is on mainboards for the CPU in the top left corner.
9 часов назад
The simplest explanation is the customer didn't follow directions and only plugged in one PCIe power cable instead of 2, and it drew 300W over a single PCIe cable, which it wasn't rated for and could've burst into flames. A fail-safe solution on retail cards is a single PSUGPU cable standard like 12VHPWR (12 pin). Another problem is cheap, undersized conductors because copper is very expensive and manufacturers sometimes undersize wires to sell cheaper cables even though the margin on cables is very high traditionally.
Thanks for the video. Do not know if this is a good suggestion or not. Solder in 90 degree pins into the hole, the solder the wires to the pins and then heat shrink.
Fact about moths: The term "computer bug" originated from a real-life insect. The first recorded instance of a bug causing a technical malfunction occurred in 1947 when engineers working on the Mark II Aiken Relay Calculator, an early computer at Harvard University, found a moth lodged in the machine's hardware.
These are better way to connect to board than soldering the connector directly to board. if connector burns and gets damaged during insertion, board does not get damaged.
🧐Not sure if I remember this correctly (melted connectors).......Looking at AC adapters... the amount of amps drawn is higher if the correct voltage cannot be sustained, so if the 12v is measuring around 11.75 then more current is drawn through the wires.....Also it may be worth noting that high ampere electrical circuits (showers) require that the core conductors be splayed to ensure maximum conductor surface contact, rather than twisting them together......I really don't think any of this relevant anyway but I thought I mention it just as food for thought 🙃.
One 8pin can handle 300w, that is why you have those daisy chained pci connectors that go from 1 8pin to 2 and the corsair 2 8pin to 16pin that can do 600w.
im pretty shure thats a eps connector that gets fed 2 pcie connections the eps/cpu connector can take alot more power then u think i just bought a nvidia cmp 100-210 mining card to crypto mine with and its basically a nvidia tesla v100 and it has this same connector with the 8 pin eps on one side and 2 pcie on the other and its working great for my card
Serious question: Can dead insects still cause serious issues within a PC? I removed *something* (looked like a long-dead moth) from an unpopulated DIMM-slot on my motherboard a few weeks back and I was experiencing BSODs before that ... a behavior which stopped after the cleaning. But I can't be certain it was the cause of the problem as I also made some changes to my BIOS re Intel C-states (disabling C10 support). So the dead moth was probably just a coincidence, but .. who knows? I also found the body-parts of a dead (and pretty toasty-looking) housefly a few days ago during a routine clean-up... not sure where exactly *that* carcass had been sitting, but it must've been either on my GPU's backplate or near its PCIe-slot. This one simply fell out while I was dusting off the card, I wasn't experiencing any GPU anomalies and wasn't looking for bugs/dead insects. With the "DIMM-bug", I was actually looking for a foreign object in there as I had reason to believe my BSODs could've been memory-related.
Nvidia RTX 20 or 30 series cards never had any issues with melting connectors and some of them were capable of pulling 300-400W of power. Why? Because they didn't have this new connector.
Being that's a full length car the owner won't be able to install it into his PC. lol I know this because we have the same card in our Dell Precision (Threadripper Pro) machine.
those connectors were always junk. I've never seen a good one the problem is the way it connects they dont have enough traction or force in the contacts, causes resistance and it goes to hell from there..🥵😁 they're just junk, even on rc cars,(the old ones,) was always a cause for batteries to not even push enough amps, just a bad design.
im curious to how this happening since i own a 4070ti (has this same type of connector with two 8 pin cables). so im starting to worry, although i dont remove my card ever. and its for sure seated all the way, i still wonder if this is just "random" or what.
plug the connector all the way in, before mounting the gpu on the pci-e slot, so you can see and check perfectly that it went full in. And never worry because a melted connector is the easiest fix ever, and is inside warranty also.
@@MrDvneil I think you missed the memo of all the gpus that have been denied from a melted connector.... Also, there is no need to plug it in before you plug into the pcie slot. You can see full well on the front of my GPU when they're slotted in. Most the melting issues I've seen just randomly happen. Thus cause for concern since this is still an on going issue. GN went over it in a vid, and a few other tech tubers too.
@@Kraxt0n no u cant see, some gpu models u can, others you can't. On Europe burnt connector is under warranty. And on na u can sent it to tony. And most burnt gpus are 4090 and few 4080, nearly none 4070
@@MrDvneil my point was, your trying to give ill sound advice on how to plug in a GPU. when no one plugs the 8pin first. not to mention i clearly stated 4070ti which are (from my searching's) visible out the side not rear of the card. you must not watch much youtube. gamers nexus, LTT, etc have all mentioned/addressed this b.s. and how manufactures deny warranty claims "customer fault". i dont live in the EU. but, turn to google to really see how well that "warranty claim" holds up. (bc it is covered, but not backed!)
@@MrDvneil ill add i am using the official cable that came with my GPU. and no, 4070s have been stated afaict but its still using that shit design. so i still wonder how these issues are arising. "are they taking their card in and out alot, just playing/running stuff on their pc, not plugging it in all the way". what...
WTF? Could have pulled the cable a bit while melted, it would bend better and maybe take it out by a millimeter.. you would not have to cut another connector
One of the Mothfets fell out. Simple fix.
clever
Hahaha, well played internets
😂😂😂 yes mothfets
Magical
Amazing
nvidia is smart tho, same thing with 4090s, they push 600 watts through a 16 pin cable while it should be a 24 pin one, they melt, and the customer needs to buy another one, amazing
But then the cables get recalled due to a fire risk..
Thank you so much for starting this video all goofy (that's the respectful way to put it)!!
I am not sure how to interpret that
@northwestrepair Easy!! Just as I said.....putting it in a respectful way!!
This is one goofy comment, that's for sure
@@NickyNiclas okay
Battery died right after the skynet joke? Too suspicious...
😵🤖
SkyNet did not approve this message :D
You monster, stop offending our wonderful overlords!
Replace your barrel jacks with XT30 connectors. Its what I'm using on my laptop now and it is 1000x better than the stupid barrel jacks that fail so often. The XT connectors are polarized so there should be no way to connect them reversed.
good idea maybe xt60s even better
@@bencromwell4453 60amps might be overdoing it. XT30 is 30 amps but probably better to downscale that to 25 amps. At 12v its more than enough for testing.
@@10100rsn30A only for instant current, the rated current is still 15A
@@bencromwell4453 There is no reason to use XT60s. It's a waste of money and a bulkier connector that will be absurd-looking with the wires running to it.
@@jorgealzate4124 yeah, I don't think he'll be pushing over 15amps on that test unit with those wires anyway.
The XT30 connectors are capable of 30amps continuous if you can fit 10awg wire on them. The wires heat up more than the connector does and even fitting 14awg wire on an XT30 to get 15 amps continuous is difficult. But I don't think he'd need that much current on the test rig. Either way, XT30 or XT60 connectors are the way.
The 150W from the spec on the 8-pin PCIe power connector is quite conservative, that is part of the reason why they don't usually melt even when not perfectly connected. But remember one of the 8 pins is used for presence detection instead of a positive, so only 3 (x2) pins are actually carrying current. When all the 8 pins are used there is a bit more headroom. Electrically, I would say they are still ultra-safe for about 225W, so this is reasonable for the TDP.
12VHWPWR is actually the opposite, it "walks on the edge", and a small manufacturing issue or improper seating can actually do damage.
Great video as always.
What does PCIe pull, like another 35W? If the max draw of the card really is 300W that's still not enough, maybe for transients. Only Nvidia would design a $5,000 card but allow some critical design flaw to be given a green light 🤦♀
Just looked at the Furmark test at the end, card is pulling 297-299W. Assuming the most robust 8-pin possible, 3 * 8A * 12V = 288W
Yep, not enough. Nvidia be goofin 🤷♂
@@The_Man_In_Red pcie slot is rated for 66w12v, and some workstation gpus use EPS 8 pin which is something like 240w
@@quantum5661 Yea according to PNY blog that's what the adapter is for, dual PCIe 8-pin to EPS-12V 8-Pin which is rated for 300W.
It's still kinda nuts to me they wouldn't just make the shroud & PCB slightly larger and use 2x 8-pin for a $5,000 card....
Maybe there's some server spec that required the cards to be under a certain size, I have no clue, but they're cutting it close either way.
I always get a PSU that's a good 100W over what I need, because its safer and keeps the PSU cooler. Pushing things to their limit tends to invite things to heat up and burn lol
Safety margins = 👍
@@The_Man_In_Red pcie connector can provide 75W according to the spec, and it indeed does since usually a couple of the phases are powered by the slot to even the load. In the case you describe 299W < 225 + 75W. As I said, opposite to the case with 12VHPWR there is a hefty safety margin in all thode numbers. Transients can be twice as that.
Nvidia does many wrong things, I agree, both in design and price. I just do not think the connector of this particular GPU is one of those.
@@Igbf just to be clear about the 75w pcie slot spec: you only actually get that number if you combine all the voltage rails, by default its only 66w sustained.
I find when routing cables, especially new, a gentle heat helps make the outer sleeve soft and this easier to work into desired path.
Based on your experience, I also found out that extreme heat liquidates the cables and allows them to be bent into any shape or form you can imagine, makes working into desired paths very easy
They're powering it by using an EPS connector, not a PEG connector. Notice the four 12v leads rather than three. Then they're using a dual PEG 8-pin to EPS adapter, using two 12v from each. The EPS connection is "rated" to 300 watts, though I'm not sure where that rating comes from since Molex Mini-Fit Jr. 8-pin dual row is rated at 10a per 12v terminal on 16ga wire for a total of 480 watts, and on 18ga it's rated for 8.5a for a total of 408 watts. Since it would be the same for the PEG connections, you still have four 12v terminals, giving you 408 watts... worst case scenario. So, overloading the connector wasn't the reason for the failure.
Also, I would like to point out that it's a common misconception that the 8-pin PEG connector is rated for 150 watts where the 6-pin is rated for 75 watts. Those are PCI Sig sensing specifications and they have nothing to do with how much current they can handle. Mini-Fit Jr. predates PCI Sig and even GPU's needing supplemental power. Hell, the first GPU's that did used the "oh so wonderful" Molex 4-pin. The 8-pin and 6-pin are actually rated the same, so with 16ga wire you're looking at 360 watts for either or 306 watts with 18ga. Even if they had used a PEG connection, they wouldn't have overloaded it, so long as they ignored the sensing specifications and use 16ga wire. With 18ga, it would be pushing it... kinda' like 12VHPWR these days.
Now Alex is going to be really pissed off at you, you're stealing his work 😂
maybe, but this channel is more like a hobby vs work. where alex worked for faster and more money while Northwest repair is a hobby.
@@erleykenneth i enjoy both but northwest repair does a better job normally Alex does other electronics too
@@PaulBlartGaming He just doing easy jobs and if you send your device and he decides to not repair it just because then you just pay alot for nothing. That's he case.
Who’s alex
@@aldeeleyquinto2741 from northridge fix.
I would have spent the wires before putting them through the solder holes so they would be at a 90° angle and you wouldn't have had to put that much torque on the solder joints.
I saw someone doing that with a circuit board and I thought it was a great idea and I kind of thought you were going to do it. All this stuff really interests me and I'm so glad I get to watch you, man.
Not a sharp 90 if anyone's going to comment but a radial 90
What I learned from the video is measure twice and cut once... well relearned that lesson? I guess I learned that a rule of thumb for construction applies to GPU repair too. But then again it is applicable for anything you need to cut to a length.
Hey there Tony, the connector on the card is actually an EPS 12V connector, which is if I recall correctly rated for 300W. (Not 100% sure about the rating) Anyway, because the connector isn't a PCIE 8 pin connector, NVIDIA includes the 2xPCIE 8 pin to EPS12V adapter.
Side Note: A lot of 8 pin PCIE cables from power supplies have quite a big safety margin in them, so they are oftentimes able to supply more than the rated 150W without problems
And the PCI express slot can deliver up to 75 watts as well. I used to have an Nvidia 1050 GPU that would get all of it's power from the slot, no cables required.
286W exactly
EPs is same as MTB cpu power i think , if you running server mtbs same as for the server GPU, I wonder why they never used the same in consumer......
@@szymontrojanowski7565how you getting 286? Google says eps12v can deliver 7 amps per circuit with 4 circuits per connector.
7x4=28
28x12V= 336 Watt
Pls do more video with that background music and edit I love them
The connector is EPS12V (300W, 4x ground 4x 12V), the same is on mainboards for the CPU in the top left corner.
The simplest explanation is the customer didn't follow directions and only plugged in one PCIe power cable instead of 2, and it drew 300W over a single PCIe cable, which it wasn't rated for and could've burst into flames. A fail-safe solution on retail cards is a single PSUGPU cable standard like 12VHPWR (12 pin).
Another problem is cheap, undersized conductors because copper is very expensive and manufacturers sometimes undersize wires to sell cheaper cables even though the margin on cables is very high traditionally.
8:33 the magic of buying two of them
The "Milling Machine". Made me laugh when I actually seen it in action with the second connector :D
Thanks for the video.
Do not know if this is a good suggestion or not. Solder in 90 degree pins into the hole, the solder the wires to the pins and then heat shrink.
Good temps for the blower type of card pulling 300W.
Fact about moths:
The term "computer bug" originated from a real-life insect. The first recorded instance of a bug causing a technical malfunction occurred in 1947 when engineers working on the Mark II Aiken Relay Calculator, an early computer at Harvard University, found a moth lodged in the machine's hardware.
you should have soldered them in a pre L bend & tinned them in a L bend before soldering them so they fit in the holes nice & tight .
1:08 Why this connector cannot be purchased ? This is the common EPS12V connector on the motherboard. A6000 uses this instead of PCIE 8 pin.
the power connector standerd is only rated for 150watt but most cables are over built to provide ~250-300watt
i love repairs of enterprise cards
These are better way to connect to board than soldering the connector directly to board. if connector burns and gets damaged during insertion, board does not get damaged.
RIP moth 😭😭
We hardly knew ye
We will always remember Dave!
The motherboard supplies some power to, my RTX 2080 super draws 47 watts from the motherboard during a 8k vr video, not much cpu gpu however
I wish they put in XT90 on those cards, also the XT60 would be the best for your little "PSU".
You should have tried heating up the short cable with the soldering iron and bent it straight and see if it would have fit
🧐Not sure if I remember this correctly (melted connectors).......Looking at AC adapters... the amount of amps drawn is higher if the correct voltage cannot be sustained, so if the 12v is measuring around 11.75 then more current is drawn through the wires.....Also it may be worth noting that high ampere electrical circuits (showers) require that the core conductors be splayed to ensure maximum conductor surface contact, rather than twisting them together......I really don't think any of this relevant anyway but I thought I mention it just as food for thought 🙃.
🔆 is way down have to push the sun into my back light, please consider😅
0:43 it's a bug!
My favourite youtuber
I just came across a youtube channel and i realized its what you were imitating in your last videos. Irs Davie504. Hes pretty funny.
report: i smack that like button repeatly until black and blue
Awesome fix trick I dig it
Why do I have the sudden urge to subscribe?
3:22 Don't forget to remember all of them in the list, because you can't bring that list with you to the past...
Been there done that... the tinned wire doesn't bend so easily.
^..^~~
"Measure once, cut twice, and the fucking thing is still too short!!"
2:08 75w pcie 8pin rated at 225w
One 8pin can handle 300w, that is why you have those daisy chained pci connectors that go from 1 8pin to 2 and the corsair 2 8pin to 16pin that can do 600w.
im pretty shure thats a eps connector that gets fed 2 pcie connections the eps/cpu connector can take alot more power then u think i just bought a nvidia cmp 100-210 mining card to crypto mine with and its basically a nvidia tesla v100 and it has this same connector with the 8 pin eps on one side and 2 pcie on the other and its working great for my card
rtx A (quadro) professional gpus use eps 8pin 12v (cpu) connector
hello fellow slapper ,,,great video ,,,I went slap happy 😁
Serious question: Can dead insects still cause serious issues within a PC? I removed *something* (looked like a long-dead moth) from an unpopulated DIMM-slot on my motherboard a few weeks back and I was experiencing BSODs before that ... a behavior which stopped after the cleaning. But I can't be certain it was the cause of the problem as I also made some changes to my BIOS re Intel C-states (disabling C10 support). So the dead moth was probably just a coincidence, but .. who knows?
I also found the body-parts of a dead (and pretty toasty-looking) housefly a few days ago during a routine clean-up... not sure where exactly *that* carcass had been sitting, but it must've been either on my GPU's backplate or near its PCIe-slot. This one simply fell out while I was dusting off the card, I wasn't experiencing any GPU anomalies and wasn't looking for bugs/dead insects. With the "DIMM-bug", I was actually looking for a foreign object in there as I had reason to believe my BSODs could've been memory-related.
would like to see how you did the milling
Hey there Tony! I was just wondering what soldering iron do you use?
Aixun
normally those cards use EPS just like cpus right?
You Da Man! .....Man!
If thy all broke down like this one the future is good 😂😂😂😂😂😂, imagine all robots short-ing out wile trying to capture you 😂😂😂😂
Nvidia RTX 20 or 30 series cards never had any issues with melting connectors and some of them were capable of pulling 300-400W of power. Why? Because they didn't have this new connector.
Can you do a Video about your test PSU with the Display that shows the power draw?
We have some of these in workstations running real-time broadcast graphics workloads 24/7, I wonder if their connectors have gone melty yet...
well i need a sneak peek into that list just in case 🙂
13:38
i already did at the start of the video 🤝
some 90° pins would have been helpful.
can you help me i have gtx 760 gpu but it's showing like another gpu tesla k10 🤔 Do you know anything about it?
any recoms for memory chip thermal pad? i need to change for my laptop , because it has thermal paste from the factory
OMG ANOTHER ONE!
Oh no your mic battery died? Time to break customers gpu and rerepair it with the mic alive this time. Lol
I learned not to cut cables too short or I have to redo all the work
people should really learn tune down there cards a little , very very few things need that much power
Great vid! and thanks for scaling down the memes!
I prefer AI alex voice and the memz
Thanks i didn't get slape from u today 😅
The graphics card failure was absolutely caused by the moth; Where do you think the expression "Bug in the system" came from? Very common.
😄
What part of this card is so expensive? where is the magic rock hidden?
3:44 that’s why I’m using the a2000. Stay off the list
the original cable seems to have had thicker wires though
Are there two guys on this channel?
nice.
"It Just works"...
My graphic card leaves the display after running for one to two minutes can anyone tell me what the problem is
Being that's a full length car the owner won't be able to install it into his PC. lol
I know this because we have the same card in our Dell Precision (Threadripper Pro) machine.
GPU has few bugs ha ?
Crimping tool and connectors..
After three more generations of Nvidia GPUs they are going to need cables the same thickness as automotive jumper cables.
came with a MOTHFET
You milled the aluminum the original connector will not fit i bet
Well I for one welcome our new AI overlords
Wrong drivers on MB and video-card can melt connectors. It happened to me.
nvidia officially sucks at properly speccing their reference cards power input sections.
those connectors were always junk. I've never seen a good one the problem is the way it connects they dont have enough traction or force in the contacts, causes resistance and it goes to hell from there..🥵😁 they're just junk, even on rc cars,(the old ones,) was always a cause for batteries to not even push enough amps, just a bad design.
Make a mod trend, replace the shit 12VHPWR with a XT60 connector
im curious to how this happening since i own a 4070ti
(has this same type of connector with two 8 pin cables).
so im starting to worry, although i dont remove my card
ever. and its for sure seated all the way, i still wonder
if this is just "random" or what.
plug the connector all the way in, before mounting the gpu on the pci-e slot, so you can see and check perfectly that it went full in. And never worry because a melted connector is the easiest fix ever, and is inside warranty also.
@@MrDvneil I think you missed the memo of all the gpus that have been denied from a melted connector.... Also, there is no need to plug it in before you plug into the pcie slot. You can see full well on the front of my GPU when they're slotted in. Most the melting issues I've seen just randomly happen. Thus cause for concern since this is still an on going issue. GN went over it in a vid, and a few other tech tubers too.
@@Kraxt0n no u cant see, some gpu models u can, others you can't. On Europe burnt connector is under warranty. And on na u can sent it to tony. And most burnt gpus are 4090 and few 4080, nearly none 4070
@@MrDvneil my point was, your trying to give ill sound advice on how to plug in a GPU. when no one plugs the 8pin first.
not to mention i clearly stated 4070ti which are (from my searching's) visible out the side not rear of the card. you must not watch much youtube. gamers nexus, LTT, etc have all mentioned/addressed this b.s. and how manufactures deny warranty claims "customer fault". i dont live in the EU. but, turn to google to really see how well that "warranty claim" holds up. (bc it is covered, but not backed!)
@@MrDvneil ill add i am using the official cable that came with my GPU. and no, 4070s have been stated afaict but its still using that shit design. so i still wonder how these issues are arising. "are they taking their card in and out alot, just playing/running stuff on their pc, not plugging it in all the way". what...
4 yellow wires? seems wrong
if the skynets come their charging ports will melt and they will kill nobody because they need humans to fix it...
WTF? Could have pulled the cable a bit while melted, it would bend better and maybe take it out by a millimeter.. you would not have to cut another connector
EPS CPU power connection to this card.
Anyone know if when you upgrade 4090 memory to 48Gb using 3090 pcb will you have 4090 with nv-link?
Highly doubtful, unless you somehow use vbios from said 3090. Anyway, how cursed the ultimate gpu nvidia won't make looks like?
Fun!
trade you
That's a $4000 video card?
I say, that's a bug.
Watching this on an A6000
gib
hey why the music gone damn it man put the music on for gpus sake
I think my 3080 runs more FPS in Timespy than this A6000 card.
What a stupid construction for the cables....
Ño more slap the like? Why?
First. hi tony
I'm showing @kawaiiguy is first, his post was over a minute, urs was 57 seconds....
Yeah just some seconds but i will give you first😂
Remember to vote people. Vote for Trump!!