Why the hate?
complicated a simple wirealso occasional fires
>>101524471It tries to burn down your expensive GPU, machine and your entire house at the same time.
>>101524471A new cable to make my box of cables even more useless
>>101524471>let's pull more power though fewer and thinner cables, what could go wrong?gee anon I dunno
>>101524471what is the little 4-pin cuck connector at the top for?
>>101524873Sense.
>>101524494>connecting a single cable as opposed to 3 awkward fat cables is more complicated>>101524502>>101524737user error
>>101524873Because everything that catches fire eventually devolves back to Berg. It is destiny. That's how we knew it would happen.
>>101524471why didn't they just use USB-C?
>>101525731usb-c can't carry 600W. also you really want something that will lock in place positively, USB is meant to be easy to remove.
>>101524471I remember somewhere kingpin said that evga want to put 2 cables on it to reduce the loadbut nvidia vetoed it
>>101525672>redesigned because... user error!sounds like a shit design to me anon
>>101524471Okay, hear me out here:>One Vcc wire (thick)>One Gnd wire (also thick)Explain why it wouldn't work (you can't)
>>101525839doesn't sound very bendy. That's the main issue with the 12 pin.
>>101525839The problem is more the contact surfaces.If you're going to push all the current down a single wire you're going to need a VERY GOOD contact surface in the connector. The 12 pin has melted because of poor contact in a single one of those connectors.
what other cable causes fires if not fully connected or bent
>>101525731>why not use something much more complicated, expensive, and lower-powerI don't understand this mindset. USB-C is literally some sort of hack that changes the standard USB 5V to a higher voltage, so more power can be transferred. It uses an awful dynamic protocol to negotiate the voltage, and all voltages but 5V are optional.
is pretty good connector with snap feedback, says 600W not 12 x nn AWG, car man looks and thinks 1.2kW starter motor cable
>>101524471Rush to the market and Nvidia literately bride its way to make it an ATX standard for their 4xxx series. The revamp 12V 2x6 is what should have been in the first place, but it would have delayed 4xxx launch. Nvidia tried pulling an Apple
>>101525927You need it to go between two nearby components and stay there forever, how bendy does it need to be?>>101525985It has been managed before.
>>101526589>It has been managed before.Yeah, I sure do want expensive, heavy duty connectors on my PC components.
>>101524471We moved from solution that worked to a solution that didn't work. How could anyone possibly be angry about that?
>>101524471new = badI bought a 4090 shortly after launch, just plug it in all the way like you would with any cable. Reddit is retarded.
>>101526002Anything that delivers power
>>101524471all you need is two single wires with a big fat screw terminal block, cables are for ninnies
let me guessyou NEED more
pretty telling nvidia is resorting to shill threads to prop up their design failure
sorry about your $1500 gpu, maybe you'll be more careful plugging in your next one
>>101529132Nvidia pressured Intel and rest of PCI-SIG to throw their experimental cable spec (limited to RTX 3090Ti reference) so they can deploy their 4xxx series at scheduled launch time. It is all done to save on PCB costs.
>>101529165I have a feeling this choice was the final straw that made evga exit the market. I'm going to keep using my 3x8-pin 3090 ftw3 until nvidia either goes back to the tried and true 8 pin, or adds a second 12vhpwr.
>3090Memory cooks itself >3090ti12vhpwr might melt>409012vhpwr even more likely to melt>Everything else Overpriced >AyymdBuggy 7xxx>IntelBuggy, too newFucking hell, where is the "just werks" GPU?!
>>101524471It's stupid and completely unnecessary, and it's also more prone to causing fires than the old connectors. Not to mention when they introduced it it's like they intentionally put it in the dumbest, most awkward places.
>>101530317Consumer GPU market is so fucked
the small sets of pins at the top capture your brain waves, processes your thoughts through the gpu, and sends them in graphical form to the glowies
where are all the user error fires for the previous connector though
>>1015303177900xtx just works for me
>>101525752Why not x2 USB-C?
>>101531383Thats because it doesn't use the shitty 12vHWP connector, trying to push 600watts through a single cable was a retarded idea im hoping pcie sig 5.1 will fix it but we will see
>>101525672It's orders of magnitude more prone to user error than 3 awkward fat cables for some reason.
>>101532272There's nothing wrong trying to do it with a single cable. The problem is that said cable was build around being small, sexy. Safety factor and tolerances were a secondary concern at best. Nvidia tried using a play from its idol, Apple.
>>101530317Ayymd 6xxx if you can still find it in stock?
>>101531223Only AD102 really needed it, the only reason it was crammed down is because Nvidia wanted cut down on PCB and power circuitry costs that would go with previous established PCIe 8-pin/6-pin setup. They also wanted to go the small, sexy route like their idol, Apple.The revised 12V 2x6 spec fixes the issues by making it impossible to supply power without it being properly seated. The problems were caused by arcing from improper seating.
>>101532287>There's nothing wrong trying to do it with a single cableYes there is even kingpin said that there should be two connectors to split the load nvidia didn't listen and here we are now
>>101531383It has that idle power bug when your screens are different sizes or something. Apparently this will never be fixed because it's been around since the 6xxx series in some form or another.>>101532304Im on Linux with a Vegan 56. I'm considering just sucking it up and dealing with the potential issues with 6xxx and 7xxx cards. Alternative is wait for a used 4070tisupper or 4090 and hope it doesn't burn my house down.
>>101532501> idle power bugI have different refresh rate screens, but no idea how to even check for idle power.My main issue is the inability to tune the fan curve. The software that claims to support this didn't the last time I checked. I want to turn 0RPM off so the fans are always running at least at low RPM.
>>101532521Me neither lol. Im guessing there's reviews or something that show evidence of it.0 rpm is cancer. Last thing I had which did that was a seasonic PSU from 2013. The fans would spin up with a WHOOOSH every 5 minutes or so. Most piss annoying thing in the world. Just fucking spin and stay spinning bastarding PSU. Returned that thing pretty fast.
>>101530317My 4090 just werks and I use it with 40°C outside. Model training on night, gayming during the day
>>101525752too bad the shitty new connector doesn't lock in place either.
>>101524471Way too much ampere on way too narrow wire.
>>101527112No no, I want my $2000 GPU to have the cheapest and flimsiest connector so it melts and I have to buy another one.>>101529132>>101529147>>101529159>>101529202
>>101527112>expensive, heavy duty connectorsI mean, yes, you're paying fucking $2000 for a video card
>>101524471Get rid of the 12V rail and replace it with 48V. Then you can use thinner wires with no problems.
>>10153397448v won't happen, due to it needing to be moron proof enough not to electrocute idiots.24v would be nice though. Double the current without changing the wires.They could make the new 12pin connector carry a 'max' of 800w, which would give a 400w overheard for shitty cables, bending, etc.
Can't hate what's not there.
>>101525672>connecting a single cable as opposed to 3 awkward fat cables is more complicatedright there you made abundantly clear you don't understand a fucking thing about electricity or cables lmao...it's 2 fat wires and a ground...lol, lmao even...you're so dumb hahahahahahahahahahahaha
>>101524471Retards got mad when they begged their mom for a 4090 and killed it because their weak hands couldn't fully seat the connector.
>>101525724>Bergcertified oldfag, I kneel.
>>101524471Try playing cyberpunk at max settings at 4k with a 4090. I wouldn't feel comfortable heating my room with such a small cable.
when are the gpus that get their power from the motherboard coming?
>>101534663
>>101534663see>>101534450
>>101534663>500W through a water-thin PCBlmao
>>101525672No, it's a design error. It'd be like if someone designed a drinking cup with bladed edges. Sure, you could drink out of that, but the average person is going to cut their mouth.
>>101524471The 12VHouseFire connector is an example of why super tight tolerances are a bad idea.
>>1015303171080ti just werks
>>101534663my 3050 low profile just werksnow I can play minecraft at 60fps
>>101524471I only really hate that it's on the side of my 4090 requiring an awkward bend.A real engineer would give it a 90 degree angle and make it reversible so you can have the cable hang down or go up whichever makes the most sense for your case.
>>101527112>Yeah, I sure do want expensive, heavy duty connectors on my PC components.Yes?
>>101537619
>>101537696A real engineer would just stop the arbitrary segmentation of server and consumer GPU design in regards to where the power connector is placed. Server GPUs have the correct idea of just not placing the power connector somewhere where it would be smooshed by the case.
Already fixed with 12V-2x6 and ATX3.1. Also a non-issue unless you bought a 4090
>>101524471Because it can easily start a fire even when properly connected.
>>101532561>Mine just werks hurp a durpCatch fire motherfucker >>101537396Based
>>101529132OEM issue
>>101534672like thats gona be any better
>>101536811If you can deliver something like 90A to the CPU through the PCB just fine, 50A to the GPU should be fine.
FTFY.
>buy Geforce latest edition n-1 ##70 card>1/4 the price of bleeding edge beta test cucks>never have any issues ever
>>101524471old pcie pwr connectors had much more contact surface area per connector. which lowers how hot it can get at high amperage. if they made 12vhpwr connectors about 20% bigger in size, there would be none of these issues besides the sense pin shit.
>>101529147melted on the end most prone to tilting out of position
