News Nvidia Seemingly Uses Revamped Power Connector for RTX 4070 FE

[Admin]

Nvidia already uses refined auxiliary PCIe 5.1 power connector on its GeForce RTX 4070 Founders Edition.

Nvidia Seemingly Uses Revamped Power Connector for RTX 4070 FE : Read more

 

[edzieba]

Since the third inescapable universal constant (after "death" and "taxes") is that the universe will always dispense a bigger idiot when the opportunity is presented: it will only be a matter of time before the 'new' connector variant melts due to poor insertion, where someone has cranked over the plug such that the sense pins mate correctly but the power pins are jammed in at an angle (regardless of what that does to the plastic housing). Nothing is foolproof to a sufficiently talented fool.

 

[Soaptrail]

The author has a sense of humor:

It seems that Nvidia is willing to take the risk with its GeForce RTX 4070 Founders Edition product,

 

[The Historical Fidelity]

Since the third inescapable universal constant (after "death" and "taxes") is that the universe will always dispense a bigger idiot when the opportunity is presented: it will only be a matter of time before the 'new' connector variant melts due to poor insertion, where someone has cranked over the plug such that the sense pins mate correctly but the power pins are jammed in at an angle (regardless of what that does to the plastic housing). Nothing is foolproof to a sufficiently talented fool.

Truer words have never been spoken

 

[Friesiansam]

Both Nvidia and AMD need to put a lot more effort into improving efficiency, whilst increasing performance. They can't carry on shovelling in more and more power, for improved performance. It's ridiculous that a top-end graphics card, can draw more than 450 Watts,

 

[doomtomb]

Both Nvidia and AMD need to put a lot more effort into improving efficiency, whilst increasing performance. They can't carry on shovelling in more and more power, for improved performance. It's ridiculous that a top-end graphics card, can draw more than 450 Watts,

Couldn't agree more. It's always going to be about Performance per Watt.

 

[NeoMorpheus]

I wonder how much "ink" would had been used if it was AMD that decided to use this power connector AND had the same melting issue.

But since its nvidia, its only a minor inconvenience...

 

[mwm2010]

Truer words have never been spoken

You are absolutely right.

 

[YSCCC]

Since the third inescapable universal constant (after "death" and "taxes") is that the universe will always dispense a bigger idiot when the opportunity is presented: it will only be a matter of time before the 'new' connector variant melts due to poor insertion, where someone has cranked over the plug such that the sense pins mate correctly but the power pins are jammed in at an angle (regardless of what that does to the plastic housing). Nothing is foolproof to a sufficiently talented fool.

Don't agree with it.

Yes user error will always happen, but when you design a card that is so tall, where 90% of cases out there will yank on the connector hard enough to render it's clipping mechanism still clipped but with a bad contact is surely a design problem. typica case width is pretty standard for decades, and when you don't specify there is a new risk, ppl would just cable manage to improve look and air flow as the old connectors being yanked similarly hard for a decade, assuming it's fine! the old 8pin don't require that bending limit or so and it don't burn here and there.

 

[The Historical Fidelity]

Couldn't agree more. It's always going to be about Performance per Watt.

Then why aren’t people buying the 4000 series? Nvidia greatly improved their performance per watt performance this gen, especially compared to AMD. People care about performance first and foremost. Efficiency is a secondary parameter to decide between similarly performing product A or B

 

[Unolocogringo]

After looking at the numbers a new 4070 is in my plans.
I can then retire my 960FTW 4gb and 1060 6gb gamer.
PPD/watt kind of thing.
That will leave me with a 1070 SC ,3060ti FE and a new 4070 FE/Asus card folding.
So some of us do care about wattage and heat produced when they run close to 24/7/365. Power outages and every three months I install windows updates shut down and dust.
Other than that they work hard folding or an occasional game on my main system.
Sorry but AMD cards use too much watts/PPD for me to consider them.

 

[cryoburner]

Both Nvidia and AMD need to put a lot more effort into improving efficiency, whilst increasing performance. They can't carry on shovelling in more and more power, for improved performance. It's ridiculous that a top-end graphics card, can draw more than 450 Watts,

The 40-series cards did significantly improve efficiency over the 30-series though. A 4070 performs very similar to a 3080, but only draws around 200 watts under load, whereas a 3080 draws upward of 320 watts. A 4080 draws around 50% more power than a 4070, and a bit less than a 3080, but tends to perform around 50% faster than those cards. Likewise, a 4090 can draw over 400 watts while gaming, but it tends to be around twice as fast as a 4070, so the efficiency level of all of those cards tends to be fairly similar, and is much better than the previous generation.

One thing to keep in mind is that there is no fixed level for what constitutes a "top-end card". A GTX 1080 Ti was $700, and drew around 250 watts, and was arguably considered "top-end", ignoring the semi-professional Titan cards that were hardly any faster, but cost a lot more. But if there is a market for higher wattage cards that deliver noticeably more performance than the more mainstream models, then it only makes sense to offer those as well.

Nvidia might shift around the model names and prices to try to convince people to spend more on a higher wattage card, but no one is requiring you to buy whatever card is at currently at the top. Niche cards in the $1000+ range are for those willing to spend extra for additional performance, who don't care much about the high power draw, heat output and pricing. If you do care, then those products are not for you, and you should look for something else instead.

 

[hotaru251]

Both Nvidia and AMD need to put a lot more effort into improving efficiency, whilst increasing performance.

this is kind of what Nvidia is doing by leaning so much into framegen.

the 4060 is "technically" a pretty efficient gpu while using frame gen for the frames u get out of it.....if you ignore downsides & extremely small support for it.

 

[newtechldtech]

I still remember that dual GPU cards maxed out at 500Watts only using two high end GPU chips ... now a single card needs more ... Nvidia and AMD are not doing their homework right anymore

 

[The Historical Fidelity]

this is kind of what Nvidia is doing by leaning so much into framegen.

the 4060 is "technically" a pretty efficient gpu while using frame gen for the frames u get out of it.....if you ignore downsides & extremely small support for it.

The efficiency graphs I saw was native resolution with all gimmicks turned off to make the numbers comparable to AMD which can’t use DLSS or frame gen. Nvidia switching from the terrible Samsung 8nm to TSMC custom 5nm (aka 4N) brought significant power saving on its own.

 

[edzieba]

Don't agree with it.

Yes user error will always happen, but when you design a card that is so tall, where 90% of cases out there will yank on the connector hard enough to render it's clipping mechanism still clipped but with a bad contact is surely a design problem. typica case width is pretty standard for decades, and when you don't specify there is a new risk, ppl would just cable manage to improve look and air flow as the old connectors being yanked similarly hard for a decade, assuming it's fine!

"The case presses on the plug" turned out to be just as much of a red herring as comparing different connector manufacturers, internal pin styles, or soldered vs. crimped wiring. All turned out to be completely unrelated to the actual root cause. Even addons (e.g. Cablemod's) specifically intended to avoid cable pressure on the card plug turned out to melt if not fully inserted, which should have been the nail in that particular coffin.

the old 8pin don't require that bending limit or so and it don't burn here and there.

A few seconds to google PCIe 8 pin melting will demonstrate that these unassailably perfect 8-pin connectors were also prone to melting when not fully inserted, just like any other power connector.

 

[YSCCC]

"The case presses on the plug" turned out to be just as much of a red herring as comparing different connector manufacturers, internal pin styles, or soldered vs. crimped wiring. All turned out to be completely unrelated to the actual root cause. Even addons (e.g. Cablemod's) specifically intended to avoid cable pressure on the card plug turned out to melt if not fully inserted, which should have been the nail in that particular coffin.
A few seconds to google PCIe 8 pin melting will demonstrate that these unassailably perfect 8-pin connectors were also prone to melting when not fully inserted, just like any other power connector.

Nope, the frequency is way up in this new connector, and the connector locking mechanism beling so loose is definitely the deisgn problem, easy proof: NV and PCISIG is now revising the pin length in the new connector to try to mitigate the severe degree of the issue.

Also back then you don't see those youtuber who do PC builds have the issue, now there are at least a few cases where the melt occurrs after months of use and claimed not touched PC, even some youtube videos showed the cable, after fully inserted, and the clip still in place, can be yanked to be meltable position, that alone shouldn't be the case, as it's the card vendors and the PCISIG in the first gen of this new connector, knows how tall the cards are, and the odds of fitting it inside a case having to pull and cable manage the card IS a design flaw.

It's just like tha Mazda Rotary engine, was it a good performing engine running really smooth? yes for sure.
Does it dies much quicker due to apex seal wear? yes for sure.
Does it need to keep burning oil to keep the engine from excessive wear? Yes
Is that a design flaw? Sure as heck it is

 

[edzieba]

Also back then you don't see those youtuber who do PC builds have the issue, now there are at least a few cases where the melt occurrs after months of use and claimed not touched PC, even some youtube videos showed the cable, after fully inserted, and the clip still in place, can be yanked to be meltable position, that alone shouldn't be the case, as it's the card vendors and the PCISIG in the first gen of this new connector, knows how tall the cards are, and the odds of fitting it inside a case having to pull and cable manage the card IS a design flaw.

"Youtuber conveniently experiences clickworthy issue months afterwards" is not exactly a ringing endorsement.

 

[YSCCC]

"Youtuber conveniently experiences clickworthy issue months afterwards" is not exactly a ringing endorsement.

neither do claiming all who suffered being idiot is.

Thing is with an untouched, finished built working fine for months, it shouldn't be buring after that, user error should be obvious in a few days at most, and with more and more power in a smaller and smaller connector the design should be more robust and fool proof, designing a connector in card costing an arm and a leg, and with the design so easy to have burnt connector due to "user error" is simply a deny of responsibility.

The 8 pin connector don't have remotely this much burnt case with the first year of release nor does it need a revision to make it safer. the standard makers are well paid to do it right to prevent this degree of error, not finger pointing to the users. hack you could try defend everything is user error, but then the sales figures will tell whether the consumers agree so.

Even Petrol and Diesel fuel gun is designed to be not able to insert into the wrong type, coz it's can result in catastropic failure of a vehicle (the product) and need to have a standard beyond barrel and funnel to prevent the user error

 

[edzieba]

Even Petrol and Diesel fuel gun is designed to be not able to insert into the wrong type, coz it's can result in catastropic failure of a vehicle (the product) and need to have a standard beyond barrel and funnel to prevent the user error

You may want to pick a better example. Petrol & Diesel filler nozzle dimensions are:
- Variable between countries (some have Petrol nozzles larger then Diesel, some have Petrol nozzles smaller than Diesel, some have them the same size)
- Variable between places within the same country (e.g. US is non-standardised)
- Variable between station types (e.g. large-bore diesel nozzles for stations primarily serving large trucks and small-bore nozzles at stations primarily serving domestic cars)
- Variable between vehicles (some have narrow filler necks, some do not)
And issues of incorrect fuelling continue to persist.

 

[YSCCC]

You may want to pick a better example. Petrol & Diesel filler nozzle dimensions are:
- Variable between countries (some have Petrol nozzles larger then Diesel, some have Petrol nozzles smaller than Diesel, some have them the same size)
- Variable between places within the same country (e.g. US is non-standardised)
- Variable between station types (e.g. large-bore diesel nozzles for stations primarily serving large trucks and small-bore nozzles at stations primarily serving domestic cars)
- Variable between vehicles (some have narrow filler necks, some do not)
And issues of incorrect fuelling continue to persist.

Nope, it's appropiate

if the variable nozzle size and you use your own method to bypass the limitation and put the wrong fuel in, you can't say the design don't remind you, but this plug design didn't, nothing reminds you that it have been slightly yanked out, and it can still supply power when yanked out partially.

In the varying coutries, petrol and diesel always don't match in size or nozzle curve, so in the same country, you can't insert into the wrong hole. for this plug, you can always plug it in and partially yanked out and not obvious unless you smell the burnt

 

[edzieba]

Nope, it's appropiate

if the variable nozzle size and you use your own method to bypass the limitation and put the wrong fuel in, you can't say the design don't remind you, but this plug design didn't, nothing reminds you that it have been slightly yanked out, and it can still supply power when yanked out partially.

There have been issues with the connector not being fully seated on insertion, not with it being pulled out after proper insertion. There is no sign the retention latch has been the issue, and even the revised-pin-length connector has not touched the latch. I am not aware of any testing that has demonstrated the plug being partially unseated just from pressure onto the cables (which would push the connector in to the receptacle).

In the varying coutries, petrol and diesel always don't match in size or nozzle curve, so in the same country, you can't insert into the wrong hole.

This is categorically false:

And, to debunk another belief held by some consumers, fuel nozzle shapes and sizes are also not an accurate indicator of the type fuel being dispensed.

The US is an example where there is no standard nozzle beyond the mandated Unleaded nozzle, leading to some states simply using that nozzle for every pump, and some implementing their own standards (which may be larger or smaller in diameter).
Over/undersized nozzles can also only prevent misfuelling in one direction - either blocking fitting diesel nozzles into petrol fill necks, or blocking petrol fill nozzles into diesel fill necks, but not both at the same time. And neither can prevent the wrong nozzle being fitted at the pump, or from a dedicated individual simply pressing the nozzle against the filler neck. Hence why misfuelling still continues to occur worldwide, even in locations where the ISO standard (petrol nozzle smaller than diesel) is used - particularly the much more destructive petrol-in-diesel-tank issue.

 

[YSCCC]

There have been issues with the connector not being fully seated on insertion, not with it being pulled out after proper insertion. There is no sign the retention latch has been the issue, and even the revised-pin-length connector has not touched the latch. I am not aware of any testing that has demonstrated the plug being partially unseated just from pressure onto the cables (which would push the connector in to the receptacle).

This is logically false, there a a few youtube videos where they DEMONSTRATED that the connector, with the latch in placed, with a bit of wiggling force, can have sides yanked out enough to the melting mark point, and ppl cable manage and pull on connectors all the time is enough of a proof that at least contribute to some of the melting cases, as your original arguement, if it can go wrong, it will eventually go wrong. Of course after it have been melted you can never determine whether it is pulled out after properly inserting the plug, or just bad insertion in the first place, the melt occurs when it is at bad contact.

and back to the fuel analogy I would not further argue with the non-sense. in most Asia countries, as well as at least some of the EU, the fuel gun angle is different, so you can't get the wrong fuel gun into the wrong hole, having some US states without this system is just means that those states have poor design or regulations which makes the fueling easier to getting stuffs damaged, not that fool proof design isn't the right thing to do.

 

[edzieba]

This is logically false, there a a few youtube videos where they DEMONSTRATED that the connector, with the latch in placed, with a bit of wiggling force, can have sides yanked out enough to the melting mark point, and ppl cable manage and pull on connectors all the time is enough of a proof that at least contribute to some of the melting cases, as your original arguement, if it can go wrong, it will eventually go wrong. Of course after it have been melted you can never determine whether it is pulled out after properly inserting the plug, or just bad insertion in the first place, the melt occurs when it is at bad contact.

Link one? A side panel pressing a connector in is a very different situation than someone grabbing a connector and pulling on and wriggling it. Side panels do not have hands.

and back to the fuel analogy I would not further argue with the non-sense. in most Asia countries, as well as at least some of the EU, the fuel gun angle is different, so you can't get the wrong fuel gun into the wrong hole, having some US states without this system is just means that those states have poor design or regulations which makes the fueling easier to getting stuffs damaged, not that fool proof design isn't the right thing to do.

The ISO standard fuel nozzle used in the EU still does not prevent misfuelling of petrol into a diesel tank. A petrol nozzle (smaller diameter, 21.3 mm) can be inserted into a diesel filler neck (larger diameter, 23.8mm) in the absence of a periphery catch device, but even with a catch fitted a petrol nozzle inserted at an angle will still trip the catch and allow you to insert it.

This situation, which is the case in multiple countries, is not the result of design for the current usage (petrol vs. diesel), but an accident of history: prior to the transition away form leaded fuels, the nozzle size was common across leaded petrol and diesel. During the transition to unleaded fuels, the unleaded petrol nozzle was enlarged to prevent insertion into vehicles intended only for leaded fuel, but diesel nozzles remained the old smaller size. Now, decades after leaded fuels were eliminated, we have a setup where you cannot fit a diesel nozzle into a petrol tank (engine will not fire, tank needs to be cleaned) but can still put petrol in a diesel tank (will destroy the injector system and possible other engine components that are fuel-lubricated), the more destructive effect.

 

[YSCCC]

Link one? A side panel pressing a connector in is a very different situation than someone grabbing a connector and pulling on and wriggling it. Side panels do not have hands.
k (will destroy the injector system and possible other engine components that are fuel-lubricated), the more destructive effect.

Gamer Nexus and Jayz did demoed it can be yanked.

no your arguement is plain wrong, able to yank it loose while the clip is in place IS a design flaw, a well designed clip would not allow that and will unclip when the yanking is hard enough to make it loose.

side panels don't have hands but there are side panels which need to slide in, squeezing close and slide sideway is similar to yanking in force, and ppl have been doing cable management all the time, which includes yanking the cables on the back panel a bit to make it not loose and float around the case. All these 100% user error BS is basically just denying any responsibility or denying that something slightly better designed could avoid.