6 minutes ago, Sethos said:

No, I don't watch JayzTwoCents trash videos and his clickbait titles.

But you post videos from Gamers Nexus instead... ok, no need to further discuss on such a level, we simply wait the release and see. Personally, I would still never connect up to four 150W PCI-E 8 Pin connectors into a single 12pin connector adapter, no matter what clickbait YT videos tell, I would not feel save seeing potentially 600W going through such an adapter.

Just now, AnkH said:

But you post videos from Gamers Nexus instead... ok, no need to further discuss on such a level, we simply wait the release and see. Personally, I would still never connect up to four 150W PCI-E 8 Pin connectors into a single 12pin connector adapter, no matter what clickbait YT videos tell, I would not feel save seeing potentially 600W going through such an adapter.

Yeah, GN is a well respected source. No chance you can argue against them. And I think few care what you'd do, because it's simply not required with an ATX 3.0 compliant PSU so it's simply not an argument until there's evidence for it. Especially when the 3000-series already has massive transient issues that could cause issues even on compliant, recommended PSUs, which has finally been improved, thus the primary issue and danger has been rectified to some extend. 

The main problem with the adapters is using it on multi-lane power supplies, overloading one circuit. The only way it would start a fire is if the power supply's overload protection failed, the power supply will just shut off before hand (usually). If you have some cheap unknown power supply, it might have faulty overload protection, but it would have to be a super cheapo piece of junk for that to be true (it's a MFR code violation too I believe in most countries, because of the fire risk).

There is almost no way a modern power supply can start on fire by overloading it, any decent PS has at least a couple methods of overload protection. In the end, it would just turn itself off.

The only other way I can think of that could cause a fire is too many amps for too thin of a cable, as some of those adapters might be too thin to handle the load. That said, I think if you buy a decent adapter it's probably fine. 

Just buy an EVGA power supply in the first place, it's not even worth risking high-dollar equipment on cheap power supplies. 

 

Edited September 26, 20223 yr by Alpine Scenery

5 minutes ago, Sethos said:

Yeah, GN is a well respected source. 

And whats about PCI-SIG as source? Respected or not? 

https://videocardz.com/newz/pci-sig-concerned-about-potential-thermal-variance-with-12vhpwr-pcie-gen5-power-cable-adapters

You would smell it burning most likely way before it caught on fire. The risk is miniscule to catch a cable on fire just because you are exceeding the rating. The ratings are way overly cautious because they have to take into account temperature variances, environmental factors, and cable age. That said, I wouldn't personally do it in case you ever left your PC running a game and then left the house, but it's still really low risk.

Edited September 26, 20223 yr by Alpine Scenery

6 minutes ago, AnkH said:

And whats about PCI-SIG as source? Respected or not? 

https://videocardz.com/newz/pci-sig-concerned-about-potential-thermal-variance-with-12vhpwr-pcie-gen5-power-cable-adapters

And still says nothing about what you are peddling. Of course there's concerns of pushing that much wattage through adapters, like there's always been with adapters. Just like there was concerns about all the transient spikes the 3000-series had on compliant PSUs with no adapters. This isn't a warning, this isn't a case study, simply a concern that should be duly noted until there's actual evidence of it not being the way forward. So your old PSU is still just fine, no need to run out and buy a new one as long as it hits the recommendations. 

People have been using GPU adapter cables for decades at this point and the 'concern' about overloading them have been there as well.

So let's fear monger when there's actual evidence for it. 

Edited September 26, 20223 yr by Sethos

Sure, and GN did some precision on the PCI-SIG statement: 

"As more details emerge, PCI-SIG may provide further updates. In the meantime, we recommend members work closely with their connector vendors and exercise due diligence in using high-power connections, particularly where safety concerns may exist.” The PDF explains that Nvidia has been testing 12VHPWR connectors to validate that prototype power supplies and cables can meet the specification of 55A continuous. During this testing, Nvidia found certain cable routing conditions led to excess heat and, in some cases, melting. The conditions required for the excess heat were either subjecting the cables to severe bending or a high number of mating cycles (about 40). Cables tested in these scenarios exhibited hot spots at roughly 2 and a half hours, and melting at 10 to 30 hours. Connectors from multiple suppliers have failed.

This is with a continuous 55A of current (or 660W at 12V), which would not be a typical load condition, especially not in gaming. Nvidia did not observe any failures on connectors with low mating cycles and without any bend."

So, seems to be a storm in the water glass pushed by some. In this case, I step back and keep quiet. Personally, I would still only buy a 4xxx card with a PSU that has a 12pin connector and not using an adapter, but that is entirely personal choice as it seems. 

Yah, I wouldn't use an adapter either, but yes the risk is incredibly low with a decent adapter, and as you say --- don't bend the thing out of shape. Power supplies aren't that expensive, better safe than sorry kind of thing.
 

6 hours ago, Sethos said:

That's not actual usage, it's just allocating that much because you have it. You see the same pattern with RAM, the more you have, the more applications with allocate. Absolutely does not mean you need 24GB of VRAM, that is insane. 

See this is what confuses me. How do we interpret these values? Just loaded into the sim. 

Dev mode fps overlay: 9.4 / 23.0

Task manager: 11.9 / 24.0

GPU-Z: 12.4

MSI afterburner: 12.4

HWiNFO: 12.4

Edited September 26, 20223 yr by Cpt_Piett

I see this topic turned in to the usual fire in a cesspit debate then. 
 

If you like the best, and run either 4K or VR the answer is simple; yes get a 3090. 
Or do as others say and ‘wait for this and wait for that and wait again’. You could wait all of your life. Life’s short, if you can afford it, get it. It’s that simple. And when the 4 series is released, simply sell the 3090. Will it have depreciated? Probably a little, bit you’ll still recoup 60% if it’s value I suspect. Just like buying a car…. 

Edited September 26, 20223 yr by Ianrivaldosmith

3 hours ago, Ponymetzger said:

I would also suggest to wait at least for the 4080 (16GB). I had a 3090 and 3090ti and these cards are not powerful enough to enjoy MSFS. 

Not powerful enough? On what planet? Of course they are. I’m getting 50-60 fps in 4K at ultra with one. Is that not enjoyable? It is for me. 

15 minutes ago, Cpt_Piett said:

See this is what confuses me. How do we interpret these values? Just loaded into the sim. 

Dev mode fps overlay: 9.4 / 23.0

Task manager: 11.9 / 24.0

GPU-Z: 12.4

MSI afterburner: 12.4

HWiNFO: 12,4

I believe the Dev Mode is displaying the actual amount in use directly by the sim.

3 hours ago, Ponymetzger said:

I would also suggest to wait at least for the 4080 (16GB). I had a 3090 and 3090ti and these cards are not powerful enough to enjoy MSFS. 
 

And again 30fps it not „smooth as silk“ or whatsoever. 

 

Totally wrong, very good cards in the wrong hands. You don't understand how a flight simulator works.

MMMmmmmm...few caustic comments on here...

 

Thought we were all on the same side, It's great we can read different opinions but try and put it better...

48 minutes ago, Cpt_Piett said:

See this is what confuses me. How do we interpret these values? Just loaded into the sim. 

Dev mode fps overlay: 9.4 / 23.0

Task manager: 11.9 / 24.0

GPU-Z: 12.4

MSI afterburner: 12.4

HWiNFO: 12,4

I think the software itself has the most accurate number of how much it allocated. However, not a lot of software will actually tell you how much of its allocation it genuinely uses, as opposed to just gobbling up a bunch of available RAM / VRAM, to also provide headroom. Fly on the exact same settings, exact same conditions on a 12GB card or 16GB card and you'll magically see it uses up much less VRAM but still works just fine.