• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Radeon RX 480 "Polaris" Launched at $199

I would simply boycott AMD, tbh. And celebrate if they get bought out.

If they release a 14nm card that is the same price/perf as a 290/390, after going on and on about the improvements they've made...

...then they deserve a boycott more than nVidia do. At least they managed to make some progress from 28nm to 16nm. Sure you'll be selling a kidney or two, but at least they aren't offering 2012 perf 4 years later.

If AMD release this and try and call it a good'n.... I will have had absolutely my fill of it and them. done/

A low power 290 or 390 or in that range is what it is, its a lower end card and ~£150 for the 8GB one.
 
i am still saying that it is more than likely 390x/980 - nano performance in the end, just based on clock and core count before hardware improvements would put it in the 390x ballpark anyway. beyond that it is a bonus.

Plus as with any presentation slide, they are only there to help the person giving the presentation. To help give a little more info to the viewer. The presentation talk is what contains all of the information. The slides only ever contain queues and small snippets of information.
I'm still expecting 'slightly better than 390X' performance myself. Or at least strongly hoping.

But I also dont have dirt under my nails trying to disprove every little negative piece of info that gets released. As far as I'm concerned, leaks that paint it in a lesser light are no more or less valid than leaks that paint it in a more positive light. But I see some of you go out of your way to dismiss negative leaks and hype up positive leaks. It's a really obvious case of confirmation bias and sometimes - downright wishful thinking.
 
Wait to see how the whole PC enthusiast internet reacts when AMD release another 290 for the same money again and call it a good'n.

You seem to have your eyes fixed on the AIB price first, then the 8gb $200+, and forget about the $200 4gb. For 1080p gaming, 4gb is more than enough, as you have argued.

Gibbo has the reference at hand, so is talking about reference, probably also about just DX11.

290? As I have already said, there is a weird language game you can play, between 290 to the 980 there really isn't anything in it, so you may as well say 'when AMD release a 980' for $200, that's pretty much what they'll do at ~110w. Is there anything wrong with that?

Can you also take your mind back to the first reveal, we all thought 390x-980, you didn't seem to be this negative back then. Pretty weird.
 
RX480 is basically a 290 with a die shrink and hardly any performance improvement.

Typical AMD under-delivering. While Nvidia is giving us 40% performance improvements on their new die shrinks and doubling the VRAM.

Which is a whole load of BS. Even if it was at 290 performance that would still be BS.
 
You seem to have your eyes fixed on the AIB price first, then the 8gb $200+, and forget about the $200 4gb. For 1080p gaming, 4gb is more than enough, as you have argued.

Gibbo has the reference at hand, so is talking about reference, probably also about just DX11.

290? As I have already said, there is a weird language game you can play, between 290 to the 980 there really isn't anything in it, so you may as well say 'when AMD release a 980' for $200, that's pretty much what they'll do at ~110w. Is there anything wrong with that?

Can you also take your mind back to the first reveal, we all thought 390x-980, you didn't seem to be this negative back then. Pretty weird.

No one wants the 4GB one and the 390 has 8GB, for £200.
 
I'm still expecting 'slightly better than 390X' performance myself. Or at least strongly hoping.

But I also dont have dirt under my nails trying to disprove every little negative piece of info that gets released. As far as I'm concerned, leaks that paint it in a lesser light are no more or less valid than leaks that paint it in a more positive light. But I see some of you go out of your way to dismiss negative leaks and hype up positive leaks. It's a really obvious case of confirmation bias and sometimes - downright wishful thinking.

The whole point is that some of the negative leaks would mean the architecture would have had to undergo some sort of serious performance regression. Which GCN4 has been stated repeatedly to have improved performance from various sources.

That steam VR score like i said lacks information to fully understand it, but would mean the RX480 (6.3 steam vr shown in the slide) would be slower than a 390 (~7 steam vr score) hence if the 2304 shaders at 1266mhz 5.8Tflops was getting less performance than 2560 at 1000mhz 5.1Tflops, then performance would have gone backwards in a dramatic way.
 
The improvements in the architecture are aimed at improving games or so i thought.
Hate to be harsh, but that is a totally meaningless/vague statement.

Artificial benchmarks should never be relied upon in terms of direct gaming performance correlation, but artificial benchmarks aren't created equal. There is no way you can assert that a card will perform better in games than in artificial benchmarks, because that doesn't mean anything.
 
Which is a whole load of BS. Even if it was at 290 performance that would still be BS.

The simple fact that we are discussing if performance is at the similar level as 290/390 is shocking considering this is on a new node process.

Compare this disaster to the Nvidia 780 to 1080 performance improvements over the last 3 years.

If they can't get a solid 30%+ performance increase on the new node, then they should just pack their bags and leave this industry.

AMD is a total disaster this last decade.
 
I would simply boycott AMD, tbh. And celebrate if they get bought out.

If they release a 14nm card that is the same price/perf as a 290/390, after going on and on about the improvements they've made...

...then they deserve a boycott more than nVidia do. At least they managed to make some progress from 28nm to 16nm. Sure you'll be selling a kidney or two, but at least they aren't offering 2012 perf 4 years later.

you know that what you are saying is nonesense ?
even if the RX 480 is only 390 performance which is not, it still took 380 price point and bumped up the performance by 40%. (assuming its only 390 not 390X or higher which could go up to 50-60%)
Nvidia took the 980 price added 50$ and added 40% performance with 1070
and they took a 980Ti added 100$ and they added 20% performance
but on your planet one is perfect and the other is the ultimate deceiver

while AMD swaps bad overclocker to apparently better ones, and Nvidia swaps good overclockers to apparently worst ones.
your bias stinks by it's lack of logic.
 
Last edited:
Only powered one up to check for noise, being to busy to test.

My comment was basically common sense, the entire internet and review sites are saying 390-390X and circa 980 levels, so I just had to step in when some people were trying to create false rumours by saying 980Ti / Titan X levels as people obviously do not realise how quick a 980Ti is.

Its a good 30-40% quicker than even an overclocked 980!!!! So in general performance I cannot see it, be good to be proven wrong as they'd sell like crazy but at same time I think people's expectations should not be so crazy or deluded. Of course in an AMD DX12 where AMD is typically stronger than NVIDIA then I suspect it could be possible in a game such as Ashes of Singularity DX12 where cards like 390X do very well, its fair to say an RX 480 will do superb.

So Gibbo is saying the whole internet are going on about 390-980 speeds and not saying they are over the mark so we can guess that's where he thinks it will be performance wise. Hopefully it's more 390x/980 though.
 
The whole point is that some of the negative leaks would mean the architecture would have had to undergo some sort of serious performance regression.
No, it wouldn't. Again, you're ignoring the possibility that this architecture is more geared towards low TDP than increasing performance.

Hell, it could turn out the new node process they're using just isn't great for high power and so this lower power direction was best suited for a new product line.

There's a whole lot of unknowns going into this new generation.
 
Back
Top Bottom