• 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 4000 Series

Wonder if any of the “basic” cards will use FE or reference design pcb so EK waterblock will fit.
Or will have to buy FE from nvidia store to get compatibility which is annoying.
 
Well seemingly for you it is as you've take multiple posts and page upon page to even attempt a semi coherent explanation of what you took issue with. All you had to say is something like, while Reflex was released at the same time as DLSS 2 Nvidia never put it under the banner of DLSS. :rolleyes:

I mean it still doesn't change that it's not a 'new' feature and it even if it is it still doesn't add up to three new features but i guess if you lack the emotionally intelligence to admit to a mistake and throwing that dead cat on the table makes you feel better then so be it.

As per your last post, it seems you are agreeing in what you said with "reflex has always been apart of DLSS before v3"..... yes or no?

Plus like i said even if I'm wrong and Reflex was not a feature of DLSS, that it was a separate feature that was release at the same time as DLSS 2, it still doesn't change the fact that it's not a 'new'

Again, simple question:

Is reflex "new" to dlss v3?

All I need is a yes or no, nothing more, nothing less.

And just to clarify.... We know it's not a new "feature" for nvidia RTX users, we know that "frame generation" is new to dlss v3 and we now know that "super resolution" is just a rebrand by nvidia for the core functionality of dlss.
 
I think the connector max's out at 600W. so the adaptor can have a different number of 6 pin inputs.

its capped at 450 watts, you'd need an atx 3.0 psu for the full load of 600w, thats what i read somewhere
meanwhile aibs can ship with 4 x 8 pins instead of the nvidia reference 12VHPWR socket for delivering 600w

As per your last post, it seems you are agreeing in what you said with "reflex has always been apart of DLSS before v3"..... yes or no?



Again, simple question:

Is reflex "new" to dlss v3?

All I need is a yes or no, nothing more, nothing less.

And just to clarify.... We know it's not a new "feature" for nvidia RTX users, we know that "frame generation" is new to dlss v3 and we now know that "super resolution" is just a rebrand by nvidia for the core functionality of dlss.

i think its impossible to reasonably comment on the nuances without knowing the actual model specs like an analytical representation of the model in an algebraic form, so everyones speculating on how those jargon's actually tie into the output but there's some support for new formats baked into the 40 series thats not available in the 30 series, so there might also be a fundamental change in how reflex gets incorporated into the rendering pipeline, though it would be interesting to read a whitepaper... that can help deconstruct these jargons
 
Last edited:
Ahhh man I don’t know what to do for my new build.
Go FE and get a water block and try doing my first loop.
Wait for an AIO card or get the Strix.
 
Good old Nvidia shenanigans.... selling what in any previous launch would have been the xx70 card as an xx80 card because of course what we really need is two xx80 cards launched around the same time where one is not only significantly shorter on memory but massively cut down elsewhere as well.

I can only imagine this is a stupid stunt to try and make the real 4080's price not look quite such a leap from last gen when they can claim '4080' pricing starts at a much lower price than it really does.
 
Last edited:
You really think OCUK and other sites will resist the temptation to incite another gold rush by opening pre-orders once they have the first stock in?

No chance, as soon as the first drops arrive in warehouses I think websites will open up preorders.
Obviously Gibbo has responded himself, but i wasnt particularly speaking for elsewhere, i fully expected them to be taking names now (or any day now), it was more a case of knowing that if OCUK was only making them available to customers on release day, you could make your mind up as late as that, and still be able to make efforts to get one, you wouldnt have been at the back of the queue.

@Gibbo Does that launch date include nvidia selling them themselves, and was someone elses earlier comment that FE's are only available direct, correct?
 
Wonder if any of the “basic” cards will use FE or reference design pcb so EK waterblock will fit.
Or will have to buy FE from nvidia store to get compatibility which is annoying.
The FE uses a unique pcb shape wth the < cut in the side so the only block that will work is the FE block from EK , i will assume like the 30 series you will also get a reference design that can be used with one block which EK will list what cards are compatible and all the rest that are custom pcb get their own blocks and they normally do it in order of how popular they are ... the Strix block was one of the 1st custom blocks they done last gen but this time they have the FE block ready to go and can be preordered now for November delivery , hopefully have a card by then :cry:
 
It's interesting that they seem to be combining Reflex with frame interpolation in DLSS3 - Reflex cuts latency by reducing frame ahead rendering I think, while frame interpolation works best when you have frames rendered ahead that you can calculate intermediary frames from and slip the new ones in the middle. So they'd seem to be mutually exclusive.

If DLSS3 frame interpolation is doing something different then who knows - if it estimates frame +1 rather than an intermediary frame then it might be acting as an alternative to render frame ahead, so you in effect enable Reflex by cutting traditional render ahead and instead use DLSS frame ahead.
 
Wonder if any of the “basic” cards will use FE or reference design pcb so EK waterblock will fit.
Or will have to buy FE from nvidia store to get compatibility which is annoying.

There's a reference design, then there's the FE design (which is unique to FE), and third-party designs. So none of the basic ones will use a FE PCB. They might use reference.
 
Last edited:
i think its impossible to reasonably comment on the nuances without knowing the actual model specs like an analytical representation of the model in an algebraic form, so everyones speculating on how those jargon's actually tie into the output but there's some support for new formats baked into the 40 series thats not available in the 30 series, so there might also be a fundamental change in how reflex gets incorporated into the rendering pipeline, though it would be interesting to read a whitepaper... that can help deconstruct these jargons

Completely agree, which is why I have also said for all we know there could be further improvements to dlss/super resolution itself outside of the reflex and frame generation feature additions.

There is no doubt nvidia have played the marketing aspect well for dlss 3 by rebranding dlss 2 to "dlss super resolution performance"

I'm hoping the DF videos will cover any changes/improvements outside of just the frame generation feature.

DLSS 3.0 is the new element, not reflex.

Strictly speaking, your question should be "Is reflex new to DLSS?"

If we're going to be pedantic...

True! :D :cry:

What is your stance on dlss 3 and it's "features"? ;) :p
 
Back
Top Bottom