Show us and maybe then we can help you ...

I would like to add that Polaris is not using TSMC's 20/16nm transistors ... 

AMD is using Samsung/GlobalFoundries 14nm LPP process node with a known die area scaling advantage of upto 15% smaller die which irons out to 17% more transistors to play with in the same die sapce ... 

Without accounting for other factors that don't contribute to performance such as the display engine or video codecs and etc a 232mm^2 14nm part can match a 546mm^2 28nm part in performance and get in range with 600mm^2 GPUs ...

Show us the page ... 

I also wouldn't get a laptop until june once the new GPUs hit the shelves ...

If AMD can manage to get a 232mm² chip beating or coming close to the monsters that are their Fiji chips (at least Fury), it would be awesome! But that could bring some trouble with cooling and heat density.

Oh, and it's a few days old but somehow I missed this piece of news:

THE PC GAMING SHOW IS BACK FOR 2016

http://www.pcgamer.com/the-pc-gaming-show-is-back-in-2016/

https://www.youtube.com/watch?v=dX_inDdvsxE

It would be nice for AMD to come close to those monsters in such a small space for their sake and I don't think heat density should be much of an issue with the clocks that GPUs run at ... 

The Fury X was a missed opportunity with AMD betting on more compute and bandwidth. Their DX11 drivers ain't bad but they would have had a better deal with a bigger frontend to lower submission overhead and 5 geometry processing units and maybe 60 CUs ? 

As for the PC Gaming Show it was very horrible from what I heard and based on the announcements last year so I'm not particularly excited for this years showing ...

I could see some issues with Football Manager provided that your running it on something ancient and likely with an Intel GMA series ... 

As for Dying Light any DX11 GPU will do since Microsoft is more strict with the conformance tests provided you also have the required video memory. The only laptop equivalents that you should avoid is the Mobility Radeon HD 5165 and below since AMD recycled some DX 10.1 parts in the product line ...

I do hope heat doesn't become a problem. After all, these cards were the "Artic Island" ones because of their focus on efficiency, but AMD will still have to come up with a good cooler and not the thing they used with the 290 series.

What I found odd about Fury was its performance disparity. At 1080p they are kind of bad performers, with the 980 coming close in many benches, and yet once resolution goes up their performance doesn't drop as much as the other cards. In any case, AMD says that GCN 4.0 brings lot of changes so we'll have to wait a bit more to see what they've done.

About the PC Gaming Show, it was something completely different than what we're used to see at E3, and at some point it was more like a late show kind of conference: one host interviewing many people in the industry with some videos of their work and some jokes.

It wasn't great or mind blowing, but it wasn't horrible either. That said, there are lots of ways to improve it even without changing the format: less talk about how great PC gaming is and more game demos running even if they are in the background would improve it a lot. Also, it would be great for all if both AMD and Nvidia took part on it and use the show to reveal some new hardware, but I can't see them both attending to the same show and not being the central point of it.

All I know is that I'll watch it again (not live as it will be too late over here).

You should figure that out for your own use cases ... 

I personally just buy for performance. If I had the money right now I would be looking to get a GTX 980 Ti but if AMD wins the next round of their GPUs with the help of DX12 to back them up a bit I might stay with the red team seeing as I'm currently fine ... 

Basically it all boils down to this, look at the benchmarks with the most performance intense or poorly optimized game/software for however you define it and then purchase your brand or whatever based on that ... 

AMD has hopes that more games will become compute limited to give them the advantage in the long run and that's exactly what their aiming to do with support for asynchronous compute shaders just like how console APIs expose support for that kind of feature ... 

I'm curious as to how much of an improvement it will be with the changes they've made to the command processor, geometry related fixed function units, and the shader instruction pre-fetching ... 

That would sound great ... 

As for me I don't do E3 so I base it on what it announced and what content they show for it ...

That's one of the reasons because I like AMD, they launch products that not only cover today's necessities, but also the ones that will come in the future. Maybe that's why their 7970/280X is still a very capable card today even thought it launched three or four years ago.

But, because everything has its cons, that forward thinking can also backfire on them. The best example that I can think of is how they had a tesselator unit years before any game used it and, when engines and games started using it, they did so with a different method that made those old cards useless for tesselation.   

I always find E3 to be an interesting time because, if nothing else, is a clear sign of how healthy is the industry.

And well, this year could be fun to watch with Nintendo having new hardware and AMD (and hopefully Nvidia) using the conference to reveal at least part of its cards and plans.