Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
Ok! Progress of a sort! But, unfortunately, your error based on your admitted "no knowledge" of the most important aspect of your own argument completely contradicts what you've been doubling and quadrupling down on for a while now.
I think I am only one of three people to point out that a purchaser negotiating with a single vendor independent of others is not price fixing. It is the kind of negotiating that must exist for the free markets the DoJ needs to enforce to in turn exist.
Exactly. I'm not one of the three but I did mention negotiation and it seems to me to be self-evident that such negotiation happens as a matter of course in all business transactions. It is the nature of business from street traders to huge corporate deals. Negotiaion is not price fixing - in some ways it is the opposite. At least that's how I see it. If anyone has ever bought from sites like Alibaba.com they will see different prices for different quantities (and these prices are still negotiable). I really don't see the point in arguing against this as it is so obvious.
Great. I give, believe any thing you want. This is just my field.
Is there a dumbed-down thread for a DAZ user with an old computer that only renders in CPU mode who is looking to buy a new gamer? computer that will work with DAZ and use gpu rendering? What is a decent Nvidia card to look for being included that will work well for DAZ with deforce clothes and hair.
10 days 17+ hours to go! I'd embed the timer, but 1) not sure how to do that and 2) might piss the moderators off...
Anyways, mountainous grain of salt but...
https://wccftech.com/nvidia-geforce-rtx-3090-flagship-ampere-gaming-graphics-card-pictured-1400-us-price/
Price subject to change at a whim of course! 24 GB for around that price would be really cool, but if that founder's card is indeed a triple slot card, consider my enthusiasm muted! I'd probably hold out for a 2 slot version... a liquid cooled single slot would be impressive, but the odds of that happening are probably less than 1%...
Why would I want a single slot? 'Cuz I want to build to that quad GPU setup, with free space left over for other non-GPU cards of course! Sure, there's PCIe ribbon cables, but those get janky...
WCCFTech also has pricing for other card ranges in another 'huge mountain of salt' rumor, but since that can change at Jensen's whim on a dime, no point getting people's hopes up. Plus it's WCCFTech, so some people won't care in any case.
So yeah, 10 Days, 17 hours+ (as of this post)...
If the power rumors are true, that'll make running 4 GPUs more of a challenge than it already is, and I'm not even sure Blender will grok two sets of two NVLinked GPUs, even if the Linux Nvidia driver does. Pretty sure I'm going to uses two systems of two GPUs each, on separate 20A circuits.
There's also a rumor(s) out there that these GPUs are being built at the Samsung fabs.
Above video is by Jim at AdoredTV, for those that might be curious.
Wow, that was a good analysis that I was actually able to follow. It supports the 400W rumor, and that means no 15A circuit could support even the GPUs alone. And the guy that wrote E-Cycles told me that he sees even an RTX2080ti intermittently pull 400W. I think 4 GPU setups are going to move into the professional-only category...
I know everyone is picking at each other about the memory and list price vs volume and Kenshaw talks about math. If Nvidia puts 24GB on the 3090 and decide to build 100,000 units that is a lot of memory. Based on a leaked PCB photo the card will have memory on the back of the board so IF it is 24GB as some rumors have stated, will have 24 1gb modules. For 100,000 cards that is 2.4 million memory modules at $13.00 a chip. That is 31.2 milliopn dollars of GDDR6X memory. I would think that if Nvidia is buying that much memory, they would get a better price than $13.00 a gig.
I'll be very curious as to how well these new cards will react to undervolting. Even so, I could live with say 2 double slot cards with my remaining PCIe slots (except the 2nd & 4th sitting under the GPUs) being available for other things like capture cards, adaptor cards, etc...
It's not like I'm going to be able to afford 4 GPUs anytime soon anyways, unless they are really cheap...
Part of me wants to rebel and embrace 64 core Threadripper with a token GPU for desktop duties. Sure, the renders will be a bit slower as they would be CPU only, but running out of memory shouldn't be an issue at that point, since I'd have at least 64 GB of memory installed in such a setup.
Dual 64 core EPYC Romes could also be interesting, but that's out of my price range.
Anyways, I'm very curious as to how Nvidia's new product stack will shake out.
And the latest rumor with image is that the 3090 is supposed to be a triple slot card, and that the special 12 pin power connector will supposedly allow up to 600 watts of power as well.. And on top of that it seems the rumor of the type of ram being used, GDDR6X or whatever it is called supposedly has heat issues.. Which means if true then this generation of Nvidia cards are well yikes in a lot of departments..
...well if that's true, I definitely will keep saving up for that Turing RTX Titan. I'd have to build a whole new system, as well as have the flat rewired, for a monster card like that.
I have used EVGA stuff for years. My last three video cards were all EVGA including my curent GTX 1070. I also now have a SuperNova 1200 PSU so I am ready for RTX 3000. I just wish that EVGA made AMD motherboards.
I hope not. If the tripple width is all on the normal fan side then okay but if they sandwich the PCB with a full width on each side, a lot of boards will not fit the card because the memory slots will interfere with the back side cooler. At least, that is how it looks on my Gigabyte board.
I have a feeling the dual sided thing is just Nvidia's Founder's Edition. There was a supposed pic of a 3rd party board with a more traditional 3 fan design. If there are no fans on the back, then the back should be clean.
The curious thing is certainly why would Nvidia do this on their card. It sounds a bit crazy. Why didn't they use 2gb stacks of VRAM instead of placing 1gb chips on both sides like that? To do that just for cooling seems odd to me. Unless that VRAM really does get crazy hot, that would be the only explanation to me. A 2gb stack of the chips may produce more heat, so they used 1gb chips and split them on each side of the board. But this raises yet more questions. If the AIBs are not doing this dual sided design, then surely the AIBs are using 2gb chips. So how well does that work?
I also cannot imagine this using quite that much power. Oh I am sure it will eat up some electricity, don't get me wrong. But 400 Watts seems too extreme, they would be pretty desperate to ship with that. They would have to be pretty concerned about AMD's competition to do that. But hey, I did tell you guys that Nvidia is hell bent on staying on top of AMD. They do not want to lose that performance crown. That is why they are launching now, and why they might just launch a crazy bonkers hot GPU. But the numbers they are talking for this 12 pin connector are its max. I think if it were to use that much, it might melt the connectors, LOL.
The hottest GPU in recent memory is in fact the Radeon 7. It hit right at 300 Watts in spite of being on 7nm. Nvidia has generally been under the 250 Watt level with their top GPUs for years. You have to go all the way back to Fermi, like Tom mentioned in his video, to find a hotter Nvidia GPU.
So I suppose the 3090 could break 300 Watts, but I don't think it will break 400. Maybe one of the AIBs do something absurd, or people overclocking, but at stock I don't see it. There is also a part of me that wonders if the 12 pin is bit of a smoke screen to throw people off.
I wrote about my EVGA experience somewhere here before, but my EVGA 1080ti died last year by lightning. They replaced it super quick considering we are on opposite sides of the country. I bought the 1080ti used off ebay, so I am not the original owner, but they treated me great. The tech I spoke to answered every question I had clearly. They will replace the product as many times as it breaks over the warranty. This stands out to me, because I do repair work, and in my field the extended warranty ends if a unit is replaced. So that is quite comforting. I actually bought that 1080ti knowing it had some warranty left, but I did not expect to use it. So I was already high on them, but that experience got me on board. And yeah, it would be great if they did some AMD stuff. At least motherboards, that shouldn't anger Nvidia.
I do not know how genuine / real the images in these articles are, but if real the 3090 is a massive card.. They compare it to a 2080 and well yes, again if real one is going to need a huge case to put it in.. Then there is what sort of strain there would be on the PCI-E slot and motherboard..
https://linustechtips.com/main/topic/1237143-the-rtx-3090-is-a-colossal-triple-slot-graphics-card/
https://videocardz.com/newz/nvidia-geforce-rtx-3090-graphics-card-pictured
One of the reasons I use a Case that can be converted into a test fig; the motherboard lies flat.
I now I have a dual slot MSI Radeon RX 570 8GB and a dual slot PNY GeFore GTX 1650 Super 4GB and only a mini-ATX motherboard in a mini-ATX case and it's very disappointing that I can't use both GPU cards in the case at the same time even though one of them would be on a PCIe x16 2.0 slot and not a PCIe x16 3.0 slot. There is not enough space. I have only 3 PCI slots and one is one of those is the very short PCIe x4 slot.
So the 3 slot design doesn't bother me. Once a expansion card is over 1 slot for my computer; 2, 3, or even 4 slots and it's all the same result: I can use only one expansion card. I could buy a single slot GPU though using a 2 slot card in the top main PCIe x16 slot and the single slot expansion card in the bottom PCIe x16 2.0 slot.
As far as power usage being 300W I guess it's not a problem after all they moved the release date from mid September or later to end of August so that's not something you do if your design is overheating and burning up computers. I don't believe the rumours of 400W as that's likely the math of overzealous GPU overclockers and not actually empirical data directly measured from the new Ampere GPUs.
It's the long-term affect on the motherboard; of course, many might upgrade before that happens, but I prefer not to have it as even an unlikely problem.
So that 12 pin connector everybody talking about is indeed real. But it is not quite what most expected. Take a look at the size of the connector compared to a standard 8 pin:
Its so cute! The 12 pins don't even take as much space as a single 8 pin does. According to the leak, this 12 pin is only on the Founder's Edition. They do claim that 3rd party boards may have 3 8 pin plugs.
Look to upgrade the PSU too
https://www.tweaktown.com/news/74663/nvidias-new-geforce-rtx-3090-will-require-huge-850w-power-supply/index.html
Not me, I got a 1000 Watts baby! I got that a couple years ago just to have extra juice for Daz multiple GPU setups.
I believe the GPUs below the 3090 are going to be more in line with their Turing counterparts in terms of power use. So the 3080 should be in range of the 2080 Wattage, and still be faster than a 2080ti. People looking at the 3070 will probably be quite happy, too. Much of this talk is about the top card, the 3090. With the 3090, Nvidia has gone all out and cranked everything to the max. This is not something they have done for many years. IMO this shows just how much Nvidia wants to keep the performance crown from AMD, like I have said several times now. I'm sure they could have shipped a GPU with less power with less performance as their top card. But they are concerned about AMD taking that crown away from them. So they increased performance as much as they could, but this performance uses more power. The 3090 is the only one that has this 12 pin connector.
The article above does say that there may be 3rd party boards with 2 8 pin connectors, so if true that would be closer to past top cards.
But a twelve pin connector is less than 2 x 8 pin, no?
If this is the case, it sounds more like a marketting ploy.
I believe they are using 1 8 pin connector and this 12 pin connector, so it is still drawing from 3 x 8 connectors off the power supply, at least that is my understanding. This could be wrong. But I did say this could be a smoke screen. Plus power supply makers would support this new format since Nvidia is the market leader. That would mean new power supplies will probably have some kind of big sticker on them promoting that they use the new Nvidia 12 pin, which is basically extra marketing.
It is going to be quite interesting how this goes. I have a hard time believing that Nvidia would really put a 400 Watt GPU to the consumer market. People are going to laugh at that, as some already are. I do think it will use more power than we are used to, but not so drastically.
I actually bought a Kill-A-Watt device recently and I have been playing around with it. My PC has two 1080tis and several hard drives, and a older i5. The 1080ti uses 250 Watts according to its testing by other sites, it could push beyond that if overclocked really hard. With Iray my system will hit 450 Watts, it never breaks past 500, and this is the entire system with two 1080tis installed. So I feel confident that a 3090 at 50 Watts more is not going to be a real problem. That would push my total system power beyond 500 Watts. Even in the more worse case situation where a 3090 hits 400 Watts, that would still only raise my system to around 600 Watts or so. And that is assuming I keep one 1080ti installed. I haven't decided if I am going to jump on a 3090, and if I do, if I'd keep a 1080ti installed. If it really is $1400 and performs like rumors say, it would be a huge upgrade.
I have a good 1200w PSU, that is under-utilised - even with the threadripper and two GPU.
If it is a genuine card, I believe it's more about marketting (and giving themselves headroom in case they need a 3090ti) than anything else - apart from the said 3090ti.
I have zero interest in a triple slot card; it's going to have to be cheap and powerful to get me to consider it.
I really like how my first gen Threadripper handles renders in Blender.
According to a new rumor the exact size is 310 mm in length and 2.75 slots for the Founders. 3rd parties could be almost anything. There very well could be a 2 slot 3rd party card with 2 x 8 pins that is not clocked as high. I wouldn't mind that myself if its close.
I do not think there will be a 3090ti, because there is no way they can go higher. Kind of like there was no Super version of the 2080ti. It is only missing like a 100 or so cores of the full die, and the 3090 may already be pushing past 300 Watts. So there is no where to go to get more performance. The only room above the 3090 is the Titan itself, which they are saying may have 48gb of VRAM. But the Titan would probably cost $3000+.
The only possible refresh would be switching nodes. If the 3090 is indeed on Samsung, a switch to TSMC's better node should provide improved performance. But these nodes are not completely compatible. To switch nodes like that would require a new design. It is possible that Nvidia has already designed these cards, but TSMC did not have enough capacity for them, that has been a big rumor for a while as to why they could be using Samsung. TSMC is supposedly going to have more capacity in 2021.
TSMC's role in this cannot be understated. They are the most important chip maker in the world now, and everybody wants chips made by them. So companies are battling it out for capacity, since TSMC can only produce so much. Nvidia took a gamble with Samsung, but may have been burned by that. AMD bought a lot of capacity from TSMC, which left Nvidia out. All this corporate musical chairs has led to the GPUs we will be seeing released soon, for better and worse.
Well, now this site is saying that the Founder's card is just using the 12 pin connector...there are no other connectors on the board. That changes things drastically for power. The 12 pin can handle up to 300 Watts, and the pcei can deliver another 75, so if the card indeed just has one 12 pin, then the absolute max power is 375 Watts. But nobody uses the max, so this card is probably just around 300 W if this is true. So that would totally be a big smoke screen, LOL. After all, why not use the old 2 x 8? I suppose that the answer might be space, as this 12 pin is so much smaller it only takes the space of a one 8 pin. That would leave more room on the board for other things. But really, it seems silly, especially if the AIBs stick to the old 2 x 8 pin designs, which is happening.
https://wccftech.com/nvidia-geforce-rtx-30-series-ampere-gaming-graphics-cards-12-pin-power-connector/
Here comes more need for higher VRAM amounts.
https://www.tweaktown.com/news/74642/flight-simulator-at-8k-requires-16gb-of-vram-geforce-rtx-3090-needed/index.html
Marketting hype; folks are talking about it and not about AMD.