• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

NVIDIA RTX 50 SERIES - Technical/General Discussion

The competitor is willing to offer me £20 compensation voucher. Imagine if I sent this back to them like that with no seal or anything.
I'm not getting anything from them other than being told to install it and if doesn't work they will fully refund it.
 
No outer box for the 5090, not even sealed. Condensation inside.


image hosting forum

Monitor box broken

Pg32ucdm *elsewhere*
5090 a competitor

Damn - that is extremely annoying and I’d be upset too.

What I’d clarify with the seller is whether the GPU was sent without packaging, as that’s pretty wild. Without a seal isn’t necessarily impossible…

@parisv did your MSI Suprim 5090 LC have a seal on the box??

As for monitors, it’s possible that this was just damaged in transit. What did the courier have to say? I’d have been horrified!
 
Good news, 5090 supply is going to be 'stupidly high soon'. Yea, let's see..... I don't buy it. Stupidly high x 0 is 0.


 
It does nothing to help with melting connectors.
Technically no as ultimately a PSU or cable manufacturer could wire things in such a way as to make the GPU think it's safe to draw more than they're capable of providing, they could in essence lie.

That's besides the point you're making though as like you say it does nothing to prevent +20 amps going down one of the 6 +12v DC lines (AFAIK the 16 gauge wire + terminal that the specs call for top out at around 15 amps each).

The newer design with shorter pin helps to reduce the chance of a bad connection and the resulting high resistance on any one pins/wires unfortunately it's still a possibility (wear & tear, manufacturing difference, etc) and if/when such a situation occurs there's no way for the device to know about it.
 
Last edited:
@SeeNoWeevil - thanks for your post. All noted.

I think the ‘most material’ bit, that @Murphy and I went on to discuss, is that in the old standard there was never a ‘fail safe’ setting in which the GPU would be told that there is ‘no power’ despite a loose connection. (i.e. in table 2 there is no ‘0w’ scenario - compare that with what is shown in table 1).

In that sense, this should in theory mean that in some scenarios where everything is ‘not in place’ in a H++ socket, the GPU is effectively told that there is no power. This is obviously a redundancy approach rather than a full solution to the problems though - it may be of limited effect as even in the H+ set-up, what wattage ought to still be limited in these scenarios

Does that make sense / do you agree? I’m learning on this subject myself, so I’m more than happy for this to be clarified if I’m misspeaking.
I think the new standard better prevents the kind of issues Gamer's Nexus replicated where the plug was pulled out considerably but purely because the sense pins are now shorter and you can't end up in a scenario where the sense pins were fully connected (thus allowing 600w) but the power pins were not and hence melting. I don't see how the Open-Open = 0w achieves anything tbh. If you could somehow get no connection on the sense pins but some power pins connected, you'd now get 0w instead of 150w. Is 150w going over 1 or 2 individual cables a problem?

I think you could sum up Buildzoid and Der8auer's comments on the sense pins as, it's not the seating of the plug that's the real problem, it's the state of the pins connecting to the GPU that's the problem and the pins do nothing to solve that.

All 50 series GPUs have the improved 12V-2x6 connector so there's no mental decision there. The only decision you have is what plugs to use on the PSU end and from my reading, the preference is you just avoid H+/H++ connectors full stop and use multiple 8pin. The reason H+/H++ cables need sense pins in the first place is because they're **** :cry:

Also, it's pretty annoying talking about all of this when there's no real agreed upon name when talking about these stupid 16pin cables. I propose everyone starts calling them 16pin Meltex.
 
Last edited:
I wouldn't be surprised to see a cable manufacture come out with a fused 12VHPWR cable, basically a 12VHPWR cable with six 15 amp car fuses.
 
I wouldn't be surprised to see a cable manufacture come out with a fused 12VHPWR cable, basically a 12VHPWR cable with six 15 amp car fuses.
Yeah, I reckon power supply manufacturers are also scrambling to make PSUs with various 12VHPWR safety features. If that happens, you can bet your ass Nvidia will wash their hands of the problem.
 
For what it is worth (sod all), my MSI 3080 Gaming X Trio came without a seal back in 2020. I googled it and posted on reddit, and everyone said that was normal.

My MSi 5080 Gaming Trio didn't have a plastic wrap seal around the box either. Is that normal? I could tell the box hadn't been opened though as it's impossible to get the cardboard tabs out without obvious signs.
 
Last edited:
I am curious though, do these DLSS haters watch films on Netflix? Or do they order the raw media every time? Do they refuse to view a jpg and insist on the raw uncompressed image data?

Intelligent compression is efficiency and scalable. Spotify streaming is very different from an LP, but the experience on average is good enough, why hate on being smarter with software to make a good experience more accessible?

Nah, Spotify sucks, flac or bust.

Good news, 5090 supply is going to be 'stupidly high soon'. Yea, let's see..... I don't buy it. Stupidly high x 0 is 0.



I doubt 'stupid high' is going to happen but surely things will improve from it being unobtanium.

For those who are going to use the octupus adapter with 4 8 pin connectors, is there a way to make it look less ugly? In my head it's going to look like a mess in the case.
 
For those who are going to use the octupus adapter with 4 8 pin connectors, is there a way to make it look less ugly? In my head it's going to look like a mess in the case.
Going to fit mine today with that cable and I got my EVGA cables out and my first thought “that’s going to be a ******* mess” so I reckon not, but I have ordered a newer cable, so it will only use it for around 6 weeks.
 
I wouldn't be surprised to see a cable manufacture come out with a fused 12VHPWR cable, basically a 12VHPWR cable with six 15 amp car fuses.
Yeah that wouldn't work. If anything it would guarantee failure. Any time there was enough of a spike to blow a fuse the rest of the cables would just draw more power till every fuse has blown. I guess it would save the card. But would need to replace the fuses regularly :P
 
Good news, 5090 supply is going to be 'stupidly high soon'. Yea, let's see..... I don't buy it. Stupidly high x 0 is 0.



Makes more sense than the current situation which is that there was 10 times more 5080's for sale than 5090 at launch, which didn't make much sense unless yield rates were terrible, which seemed weird given a mature node. Nvidia is hiding them 5090 chips from board partners lol
 
Last edited:
Much easier to make up gaps in the data you have then to predict the future
That's not how FG works at all, though. It's predicting nothing. As Nvidia themselves, together with many reviewers said, FG only compares 2 GPU generated images (first and second frame) and then generated frames in between. There's no predicting anything at all involved in that, it's just trying to very quickly generate something in between, often with very bad quality (because it has to be very quickly done).
 
Jensen wants exactly that. He wants to be like Apple and exclusively manufacture all their own hardware.
Which is exactly why AIB for years already are being pushed by Nvidia as the bad guys with very restrictive contracts and high pricing on components, so they always will be more expensive with very little freedom to actually change anything in the spec.
 
Back
Top Bottom