Very interesting, but then why does the water cooled seahawk offer only 200w? And what overclocking tools did they use to get 291w?
I'm trying to wrap my head around why anyone would care about power consumption when discussing overclocking.
The very nature of overclocking causes power usage to increase exponentially as an after effect. And anyone overclocking anything understands this.
With some video cards and some overclocks they can end up using a little more than double original power after the OC.
It's just part of the game.
Despite AMD is telling the world that the $649.99 Geforce 1080 is too expensive, Nvidia partners insist they can sell as many cards as Nvidia can supply. Now there is a faster version coming, based on GP102 even bigger graphics chip. We don’t have many details about GP102 other than it is fast enough for Nvidia to call it Titan. There is a chance that there will be a Geforce GTX 1080 Ti branding but our sources are telling us that Titan is more likely.
Boutique vendors including Origin PC, Maingear, Falcon Northwest, Velocity micro and a few others tell us they cannot get enough of the high end cards, as this is what their customers want. Now with the fade out of the original Titan and the fact that Geforce GTX 1080 is faster than the original Titan card, there is a space for a new card.
We expect that the card will sell for $999 and people are willing to pay for it. Our sources are confident that this is a really big chip, much bigger than the GP104 with its 314 square mm and yet smaller than gigantic GP100 used for Pascal supercomputer cards. This is a bigger Pascal based chip, manufactured using TSMC's 16nm FinFET process with more Cuda cores enabled compared to the GP104 and Geforce GTX 1080.
This is not a HBM 2.0 card, as far as we know, the card will use GDDR5X memory. If it does even a bit better than the Geforce 1080, which was the most successful high end selling product in more than a decade, then Nvidia will be laughing all the way to the bank.
The card has its clock default frequencies set at 1898 MHz (boost) / 1759 MHz (base) with a reference clocked 8192 MB GDDR5X / 10108 MHz effective data-rate on the memory. With the help of their tweaking software by pressing an OC mode you can instantly get another boost in perf as the clocks will change to a GPU Boost Clock : 1936 MHz , GPU Base Clock : 1784 MHz, that however requires the software to be run in the background at all times.
I'm puzzled. Why those two modes with a difference of 38 MHz (OC) and 25 MHz (base)? What is even the point in that miniscule difference? Why did Asus bother to add it?
What makes you think that next year's generation will have better OC's? It is in NV's interest to keep the same steady performance progression and that includes OC capabilities.
What makes you think that next year's generation will have better OC's? It is in NV's interest to keep the same steady performance progression and that includes OC capabilities.
Nvidia gimps the OCability of the mid range cards so that they won't step on the toes of their over priced high end chips. The only down side is that the new mid range cards don't look very good compared to last gen's OC'd high end cards, but when you are trying to rip off your entire customer base twice each generation, something has to give a little and the 1070/1080's poor OCability is what had to give, same as with the 600/700 series.
Nvidia gimps the OCability of the mid range cards so that they won't step on the toes of their over priced high end chips. The only down side is that the new mid range cards don't look very good compared to last gen's OC'd high end cards, but when you are trying to rip off your entire customer base twice each generation, something has to give a little and the 1070/1080's poor OCability is what had to give, same as with the 600/700 series.
Part of it is planned obsolescence. It forces people to buy a higher tier of card if they want the higher performance.. instead of buying a lower tier card and overclocking it to the sky and getting the same performance as the higher cards for less money.
Sure, they gimp them, but it sounded like you didn't think they would equally gimp their "high end" cards either. Probably just me misunderstanding your statement.
I predict they will gimp every generation forward to make performance end up exactly where they want it to be. I don't really make a difference between what you guys call "mid range" and "high end" any more. We are getting a new top card every year, and it is perfectly tuned to improve by 20-30% without OC'ing "too much". That's it. Doesn't really matter what the properties or the size of the chip is anymore.
You can still do that though! Just needs a bit of patience
Received my Zotac GTX 1080 FE this afternoon and spent time putting on the EK GTX 1080 waterblock. All is running well. Fast card even at stock. I'll be working on overclocking over the week end.
And some soldering skills :\
Asustek Computer, viewing that gaming notebooks have become a main source of consolidated operating profit, has recently strengthened marketing strategies for the second half through adopting Nvidia Pascal-architecture GPUs including GeForce GTX 1080/1070/1060 to hike hardware specifications as well as setting retail prices comparable to those quoted by Micro-Star International (MSI) in the US and Europe and those set by Haier in the China market, according to industry sources.
Was up way too late last night playing with it. I'll get back to you later tonight with the answer.What's the max power target?
Max OCed Strix almost matches stock Geforce GTX 1080.
:thumbsdown::thumbsdown::thumbsdown:
Oh wait. That was the 1070?
:thumbsup::thumbsup::thumbsup:
Same power level as the 1080FE, too...