I think there's a chance they forget the 6pin connector, if not on the 1060 the 1050 will surely lose it, at least for their reference-founders model, oems will include double 8pin overkill as always so no worries

I would like to know specs on this, thought i think might be asking to much if these use GDDR5X or even HBM2 which was the whole point of me waiting for pascal, cause i anit pay more them 300$ for gpu

970-980 performance at 200 usd?

i think its better to wait for Ti version if it exits in future or jump to 1070.

This is the card I'm most likely to be interested in...if the price and performance is right.

Let's hope it wont be a another disappointment like gtx960.

It needs to be priced around $200 to $230 and have at least GTX980 performance to avoid being a let down.

It needs to be priced around $200 to $230 and have at least GTX980 performance to avoid being a let down.

From 970 you'll probably need to look at 1070 atleast if you want a real upgrade. Your card (if oc'd) is already at 980 level.

I would like to know specs on this, thought i think might be asking to much if these use GDDR5X or even HBM2 which was the whole point of me waiting for pascal, cause i anit pay more them 300$ for gpu

It usually releases at about $240-250ish. The GTX_60 series aims for value. Often the best bang for your buck (from nvidia camp).

Is it just me who thinks the form looks like it was created in MS Paint? Especially the top section, the turquoise clickable links above the ad choices?

From 970 you'll probably need to look at 1070 atleast if you want a real upgrade. Your card (if oc'd) is already at 980 level.

I'm always throwing new computers together that you guys don't know about. At this point, it's unlikely that my 970 is going anywhere before Volta launches.

970-980 performance at 200 usd?

That would most certainly deserve my money.

I'm always throwing new computers together that you guys don't know about. At this point, it's unlikely that my 970 is going anywhere before Volta launches.

i Believe the same, 1080 is underwhelming, maybe polaris will be something...

i Believe the same, 1080 is underwhelming, maybe polaris will be something...

I think it is anything BUT underwhelming.. and that some people are just crazy but hey to each there own

These mid-perf 680/980/1080, now so called "high-end" have never been anything special. They became "high-end" & priced accordingly since Kepler and still get away with it, now at even more absurd price

-How to make even more money; 
some nv troll: what if we sell stock as a founders edition, its sounds more "prestige" 
this way we can charge extra, 
we will thrown in that vapor chamber bs and all that., 
problem solved 
free profit

Maybe Polaris will do what 5850/5870 & 6950/6970 did back in Fermi days, NV couldn't rely on 460/560GTX to sell them that high, they weren't "strong" enough. Imo, The only interesting thing about Pascal is full GP100 chip, a upgrade itch @ 5-20 fps difference e.g. 980TI vs 1080 is just waste of money for 200€ more, I mean by GP100 it will be 40-60fps, now that's a itch for sure 😀 I personally will try to skip full GP100 Pascal and get Volta or by Amd Navi. Hopefully, well I know those two will be something else, what 10nm? 🤓

It all started with that HD 7970 and GTX 680 era heh

This is the card I'm most likely to be interested in...if the price and performance is right.

Even if it's reference 980 performance that's pretty close to what our 970s perform at even without their OCs that are beyond what 980s can reach. Not a worthwhile upgrade at all.

These mid-perf 680/980/1080, now so called "high-end" have never been anything special. They became "high-end" & priced accordingly since Kepler and still get away with it, now at even more absurd price

That's what I said the day nVidia originally revealed they'd be selling their mid range chips as high end and saving the real high end to become the Titan. Everyone's creaming their pants over this 1080. It should be called the 1060. The 1080 Ti is what the 1070 should have been, and whatever their upcoming Titan card is should have been the 1080. I remember the days when you bought an x80 you got the full chip, not an architecture below, not a cut down chip, the full blown thing. The 1070/1080 seem great because we've been fed crap for so many years at such a high price that any performance gain for a seemingly non-insane price seems great. This is what happens when there's no competition or little competition that avoid each other with specific release cycles. When was the last time an x80 released could truly max out any newly released game of it's era at a standard gaming resolution like 1440? I'm talking about completely max.

IIRC, the midrange of the past were supposed to go on par or faster than the previous high ends (6600GT and 7600GT) Incredible value that time, those times are long gone T__T Then the 8600GT/GTS were mocked because its more like 7600-7800 performance then when shader heavy enabled games came out it finally went toe to toe with the 7900s

The thing is... Why should Nvidia release anything more powerful than the 1080 when there's nothing to compete with it?

Even if it's reference 980 performance that's pretty close to what our 970s perform at even without their OCs that are beyond what 980s can reach. Not a worthwhile upgrade at all. That's what I said the day nVidia originally revealed they'd be selling their mid range chips as high end and saving the real high end to become the Titan. Everyone's creaming their pants over this 1080. It should be called the 1060. The 1080 Ti is what the 1070 should have been, and whatever their upcoming Titan card is should have been the 1080. I remember the days when you bought an x80 you got the full chip, not an architecture below, not a cut down chip, the full blown thing. The 1070/1080 seem great because we've been fed crap for so many years at such a high price that any performance gain for a seemingly non-insane price seems great. This is what happens when there's no competition or little competition that avoid each other with specific release cycles. When was the last time an x80 released could truly max out any newly released game of it's era at a standard gaming resolution like 1440? I'm talking about completely max.

That Era could have been only due to MADDx5/VLIW5 moving into denser node and gaining big on SP/TMU/ROPs: Chip: SP/TMU/ROP/Clock (node) HD 2900: 320/16/16/600MHz (80nm) HD 3870: 320/16/16/775MHz (55nm) HD 4870: 800/40/16/750MHz (55nm) HD 5870: 1600/80/32/850MHz (40nm) ... Then there was practically downgrade generation of HD 6870 as it was cut down in comparison to HD5870, and VLIW4 (both still 40nm) failure of 6970 which had performance advantage in simpler loads and choked a bit more in heavier than VLIW5... HD 7970: 2048/120/32 925MHz (28nm) Therefore I would not say there were some drastic architectural improvements allowing that mentioned jump in performance. I would say that source was denser manufacturing node allowing for bigger, more robust chips. Chips practically more than doubled in performance per generation and each of those generations lasted around one year. That's something which outpaced game development cycle. We may see similar thing now. (but not yet as Pascal is underwhelming for not using node advantage and Polaris is entry/mainstream targeted) As we did sit overtime on 28nm, now finally skipped 20nm for 14/16nm. We'll see some substantial boost in performance over what we have if it gets fully used. But both nodes are expected to be much shorter lived than 28nm as 10nm is in production and 7nm based on 10nm is behind corner for 2018. Historical trends show that this 15% GTX 1080 has over 980Ti will be short lived. As prices will be. It will start with Vega which will force nVidia's hand. And then it will go fast. 2017 may be over pretty soon and we may see cards delivering 3 times as much performance as GTX 980 Ti/Fury X. And considering that GP100 confirmed that 300W GPU can be done on 16nm, I would not settle for 15% 'upgrade' over last generation as it is not using 16nm node size advantage at all (other than passive access to higher speed and lower needed voltage). This actually leads me to funny thing. Some people claim that this move from 1450MHz to 1750MHz on Maxwell to Pascal is not due to move to 16nm, but because Pascal is better architecture. 16nm costs bit more per transistor than 28nm. GTX 1080 has less transistors than GTX 980Ti (good compensation for manufacturing price). But then it would be smarter to make this quite small Pascal on 28nm, decrease density a bit over GTX 980 Ti. Get same clock as 16nm achieved. (Just with bit higher power consumption which would still be better than what AMD has at house due t even lower density.) But that's the thing, it is 16nm which allows for higher clock, not Pascal's architecture.