When will the standard GTX 1080 release?

Apr 10, 2015
128
0
10,710
I'm a bit confused over the GTX 1080 founders edition and it's release. Do the cards release on the same date? And what's up with MSI ect making their own version of the founders edition. Will It be worth the extra money buying the founder's edition? Does the card have any benefits at all?
 
Solution
Custom cards will likely follow the FE release by about a month. It really depends on how long the AiB's have had with the 1080 silicon. It takes awhile to design custom coolers and PCB's and proof them.

As for the AiB's with the FE cards, they aren't making their own. This is essentially the same as the reference cards of old. It comes down to the AiB's not having custom cards on Day 1, so they take the reference cards from nVidia (Foxconn has traditionally been the manufacturer) and they slap their own stickers on them. They didn't build or design them. It's done this way so there are cards on Day 1 and it's not like AMD's paper launches.

As for benefits, you can buy one right away. Beyond that, nVidia puts a lot of time...
Custom cards will likely follow the FE release by about a month. It really depends on how long the AiB's have had with the 1080 silicon. It takes awhile to design custom coolers and PCB's and proof them.

As for the AiB's with the FE cards, they aren't making their own. This is essentially the same as the reference cards of old. It comes down to the AiB's not having custom cards on Day 1, so they take the reference cards from nVidia (Foxconn has traditionally been the manufacturer) and they slap their own stickers on them. They didn't build or design them. It's done this way so there are cards on Day 1 and it's not like AMD's paper launches.

As for benefits, you can buy one right away. Beyond that, nVidia puts a lot of time and effort into their reference designs. In the beginning, they are the ones with the most experience designing a PCB and cooler for the GPU. They also tend to design a fairly robust power delivery (VRM) system for their cards. The other attractive part of a reference card is there are usually waterblocks available on Day 1 and are typically the most compatible with third-party coolers during the lifecyle of the GPU. So if you have a custom loop, reference cards are very attractive.
 
Solution
The one thing that is different about this launch and the FE cards is that nVidia didn't really sell the reference cards directly. It was always done via their AiB's. I'm not entirely sure why nVidia has decided to sell them on their own this time though. As for the price premium it's been debated as an "early adopters tax" which might be partially correct. However it also maybe to not directly compete with their AiB's and piss them off. The likes or MSI, EVGA, ASUS, etc have stuck with nVidia and it would be very disloyal to undercut / compete directly with them now.
 

zornyan

Reputable
Dec 4, 2015
97
0
4,640
the main benefits of the non reference designs are.

better cooling, blowers aren't great imo, and are generally loud compared to aftermarket fans provided by the likes of asus, msi etc.

better overclocking, msi gaming editions and various others will/do have a second power connector, it was mentioned by several testers that the single 8pin was not really enough to get decent overclocking headroom, adding a second power supply will up how high you'll be able to hit.

generally the non reference cards overclocking better than the reference designs, due to the better cooling and power solutions etc, as well as being cheaper.
 


I agree with some points and wholeheartedly disagree with others. Generally speaking the custom cooling designs are better in single GPU systems overall. Their main problem is in a single GPU system, they tend to raise ambient temps in the case because they don't exhaust all the hot air outside the case. In multi-GPU systems, the third party coolers are even more susceptible to air starving when the spacing is close (top card typically gets hotter).

I think that the addition of another PCI-E power connector "could" enhance overclocking, but some of this will be dictated by Turbo Boost 3.0. nVidia has strict control over the Power Limit. So even though the card may have thermal headroom (provided by the custom coolers), Turbo Boost will only allow an additional 12% on the Power Limit. Of course nVidia could loosen this in time. I'm also sure that there will be some custom / modified BIOSes to circumvent the Power Limit. For now though, 12% above the nominal TDP is all that is allowed.

As for power regulation, don't sell nVidia's design short. They put a great deal of effort into the VRM's on their "reference" / FE boards. Their PCB's are designed for stability as well. I think it speaks to how well nVidia's design is that it is such a good overclocker. Most reviews I've seen show them consistently hitting over 2.0GHz when the fan is cranked up. Certainly most users are not going to run their fans at 100%, but with custom cooling (or even better watercooling) they should have no problem hitting these speeds with an acceptable noise profile. As I postulated in my earlier post, there have already been announcements for watercooling blocks for the 1080. I expect that most of them will be available now or very shortly. So this gives the FE boards a leg up on the custom 1080's (if they have a redesigned PCB) for those people wanting to watercool them right away.

As for price, this is definitely something I can agree with. They will be cheaper and that is nVidia's intention. They don't want to be called out by their AiB's for competing with them directly. I think this is a smart move by nVidia, their AiB's have been loyal to them, and nVidia would be foolish to disrespect them now.