Nvidia GeForce GTX 1080 (16nm Pascal)

[Tarinder]

Blimey,Nvidia kept things close to their chest until the last minute. Tarinder should have a beer from all of us for that(assuming he drinks of course).

Is the Pope Catholic?

[CAT-THE-FIFTH]

Is the Pope Catholic?

I am not sure at times.

[Corky34]

Wow, nVidia must have given out a lot more cards than we know about if all those forum users are making comments about the PCB etc.

So many of them in fact that i couldn't find a single post in that thread from a single person whose even seen one.

[kompukare]

And they said that they spent billions on this? Not USD for sure, some poor country currency maybe.

That was always hype.
There is no way they spent $billionS on Pascal unless they include every bit of R&D since Ferni or something.
Nvidia's total R&D figures are public at about $300 million a quarter. And not all of that would have been spent on Pascal.

[Jedibeeftrix]

Looking at an unopened Gigabyte R9 390X G1 that I picked up for £250 (standard price for an R9 390X is £330-£360 in UK money).

This is getting 50-66 percent of the framerate of the GTX 1080, but for slightly less than half the price ($599 for the non-Founders edition translates to roughly £500 inc VAT).

Knowing what we know no about likely performance of upcoming 14/16nm products, should i be sending it back?

[kalniel]

Knowing what we know no about likely performance of upcoming 14/16nm products, should i be sending it back?

Q1: Does it perform as fast as you need for your budget?

Q2: There is no question two - new cards won't make your existing card go slower out of envy.

[kompukare]

Wow, nVidia must have given out a lot more cards than we know about if all those forum users are making comments about the PCB etc.

Surely knowledgeable people would be able to judge these things from studying the high res photos from the TPU review. Or the hints from reviewed like computerbase saying they suspected their overclock attempts were limited by the card PCB.
Anyway, all of this shows the F Edition is not worth $100 extra. I think that was always a given, buy some people succumbed to the hype and were expecting especially binned chips and ram.

You're disappointed because it's delivered the expected gains, with a drop in power consumption to boot? What were you hoping for?

I suspect darcotech might have been expecting IPC improvements whereas Pascal send to rely on clock speed only. Of course, there clock speeds are not by accident so it seems Nvidia spent time optimising for those. And since the card's power usage of the same as the 980 it must be running near 16FF's node optimum.

[Jedibeeftrix]

Q1: Does it perform as fast as you need for your budget?

Q2: There is no question two - new cards won't make your existing card go slower out of envy.

It's for the folks (Skyrim and Rome2). Build is nearly done, and can't be delivered until mid-june.

Today the price performance is great, will it still be a good purchase next month given it will also be hotter and noisier than a 14nm polaris...?

390X is unopened, could go back.

[jigger]

Looking at an unopened Gigabyte R9 390X G1 that I picked up for £250 (standard price for an R9 390X is £330-£360 in UK money).

This is getting 50-66 percent of the framerate of the GTX 1080, but for slightly less than half the price ($599 for the non-Founders edition translates to roughly £500 inc VAT).

Knowing what we know no about likely performance of upcoming 14/16nm products, should i be sending it back?

It will probably be more like £550-£630 for the first few months.

[jigger]

I would have liked more performance for this kind of money even if that meant running hotter and pulling more power. It's all a bit tame and middle of the road for me.

[Tarinder]

I would have liked more performance for this kind of money even if that meant running hotter and pulling more power. It's all a bit tame and middle of the road for me.

This is what happens when real competition, for the time being at least, is absent from the high-end graphics space.

[kalniel]

Surely knowledgeable people would be able to judge these things from studying the high res photos from the TPU review. Or the hints from reviewed like computerbase saying they suspected their overclock attempts were limited by the card PCB.
Anyway, all of this shows the F Edition is not worth $100 extra. I think that was always a given, buy some people succumbed to the hype and were expecting especially binned chips and ram.

I think armchair knowledge is quite different from real reviewer knowledge But in any case, the sales will speak for themselves to determine worth - if enough people are willing to pay a premium for early access (hint, in just about everything else to do with computers, they are) then it'll have objectively been worth the extra and we can't really argue otherwise. On the other hand if nVidia sit on a pile of unsold cards then it won't have been. I have no way of knowing the numbers in either case, but I think nVidia have smart enough guys when it comes to assessing the market to get it right.

I suspect darcotech might have been expecting IPC improvements whereas Pascal send to rely on clock speed only. Of course, there clock speeds are not by accident so it seems Nvidia spent time optimising for those. And since the card's power usage of the same as the 980 it must be running near 16FF's node optimum.

Fair enough, but graphics cards are almost never about IPC improvements - nonetheless there are some small bits in Pascal, such as the latest mem compression etc. But yes I think it's safe to say that if they manage this kind of perf/watt increase without significant IPC then it was still worth taking the route they did - the late tick-tock Intel strategy was much the same - spend one generation optimising for a node transition and another adding new things for the same node. Trying to do both at the same time adds additional delays which given we've waited so long on 28nm aren't really wanted!

It's for the folks (Skyrim and Rome2). Build is nearly done, and can't be delivered until mid-june.

Today the price performance is great, will it still be a good purchase next month given it will also be hotter and noisier than a 14nm polaris...?

390X is unopened, could go back.

I'd stick with what you have personally - mid-june is still too early to take advantage of new chips - what little there is available won't be easily within budget and you'll have to pay the early adopter tax. AMD cards are especially forward looking so you haven't made a bad call and if the performance is good enough now it'll still be good enough when the new cards come out.

[MrRockliffe]

There's nothing stopping you from underclocking and undervolting a 1080 You could underclock it to 980 Ti performance and save a fair bit of power.

[krazy_olie]

There's nothing stopping you from underclocking and undervolting a 1080 You could underclock it to 980 Ti performance and save a fair bit of power.

That can cause instability too but one presumes there will be some window for doing that. With the variable frequencies, boost and all that just capping your frame rate can have similar effects without having to faff around

[darcotech]

I suspect darcotech might have been expecting IPC improvements whereas Pascal send to rely on clock speed only. Of course, there clock speeds are not by accident so it seems Nvidia spent time optimising for those. And since the card's power usage of the same as the 980 it must be running near 16FF's node optimum.

Fair enough, but graphics cards are almost never about IPC improvements - nonetheless there are some small bits in Pascal, such as the latest mem compression etc. But yes I think it's safe to say that if they manage this kind of perf/watt increase without significant IPC then it was still worth taking the route they did - the late tick-tock Intel strategy was much the same - spend one generation optimising for a node transition and another adding new things for the same node. Trying to do both at the same time adds additional delays which given we've waited so long on 28nm aren't really wanted!

First of all, this card is GTX 980 successor. So it should be compared to it.

Hi I am actually referencing to NVidia presentation of this card where they explicitly said they weren't just gunning to enjoy the performance gain from new node. They actually invested billions/ "several thousands of engineers worked for several years " to make even more. I do not see that. Do you?

You understand that gain they made comes from new node, no extra as they told.
Most of other thing presented were software solutions (where they are stronger than AMD), but nothing breathtaking.

We need higher performance on high end in order to have great performance at middle class, the one sells the most to gamers, and to have higher baseline.

They probably have GTX 1080 ti version ready, just wait for AMD to see how good (or bad they will be).
If they weak you will pay 1080 599USD for sometime. If AMD comes strong (at the end of the year) then nVidia will come up with 1080 Ti for that price.

So I will wait.

[HalloweenJack]

3dmark vr? or any DX12 titles??