AMD Polaris 10 GPU Reportedly Offers Near 980 Ti Performance For 300 USD

[Guest]

8 hours ago, IMPERIUS said:

I hope thats true

 

edit: the 300- series was also apparently going to "destroy the 900- series from nvidia" tho that didn't really happen...

Noticeably better, usually $20 less. It's better but they didn't have enough advertising to convince the masses that the 970 was lesser. And, of course, they launched a year later.

[thekeemo]

51 minutes ago, branden_lucero said:

all Ti and X cards should no longer exist. they're barely worth a fuckin' thing over their non-Ti and X counter parts. 

uhh what? The higher end the card is the more price gouged it is.. this is how it always has been.

[ApolloFury]

Isn't Vega suppose to be the higher-end cards?

[ApolloFury]

10 hours ago, IMPERIUS said:

I hope thats true

 

edit: the 300- series was also apparently going to "destroy the 900- series from nvidia" tho that didn't really happen...

Well I wouldn't recommend the GTX 970 over the R9 390 and the GTX 980 is almost a tie or lost with the R9 390X at 1440p or above in recent games, even in Witcher 3.

4:21 of this video.

The one downside is power consumption and heat, it is an older architecture afterall.

[unsubscribed]

First of all it definitely will not be $300, more like $400. It did cost AMD a small fortune to produce these products and they will charge at least that much, especially since they can as we all know NVidia is not going to drop the prices to these levels either. Not for new GPUs.

On the same note by the time it reaches UK it will cost around £380 meaning around €500 in Europe (translates to around $550). So that's that.

 

As for performance differences between AMD and NVidia to me it is ridiculous. If you take same tier GPU from both and slam it into two separate rigs bearing same configs no one would be able to tell which is which by only using them to play games (I know I know, there are few titles out there customized more for one or the other).

 

So all in all it is futile argument who does best. Unless one of the companies snoozes or does a major screw up it does not matter in the end. Go with the GPU you like and can afford. The rest is (sorry for my language) just a pi55 contest.

[Citadelen]

9 hours ago, djdwosk97 said:

Well, there have been tens of millions of consoles sold, and AMD supplied the hardware for all of them, yet AMD is still I serious financial trouble. Clearly consoles aren't making a big impact, which leads me to the conclusion that they're not that profitable for AMD. Which makes sense since they had to undercut Nvidia in order to get the console contracts.

Nvidia didn't have the expertise to take up the contract, nor did they have the manufacturing capacity.

[Thony]

Maybe it does water-cooled and overclocked. 

 

Out of the box I highly doubt it.

[Paragon_X]

If we accept that AMD can pull that off there are some reasons i could see that this could happen.

1) smaller Dies = smaller cost = cheaper to sell while keeping some good margins
2) Market Penetration = AMD's market % is really low, if they can push to get more people to buy AMD cards they can push their technologies and contracts further
3) Cont from point 2 strategic alliances. More deals with developers if they have bigger market share. 
4) keep investors happy with big numbers

[Tomsen]

Have my doubts regarding the speculated prices (~$300), especially at launch. Perhaps thats the targeting pricerange AMD is going for, but doubt we will see them until vega appears. I don't see that polaris will coexist with the current fury lineup, not going to make sense. Fury lineup is expensive to manufacture due to HBM and the extreme huge interposers, and it seems they will comptete in the same performance-space. I'll expect AMD to make sure they empty their current lineup (comes in the form of good deals), to prepare for polaris launch.

 

The performance metric, I'm not so surprised by. AMD have already showcased both polaris 11/10 running atleast 1080p/1440p at 60fp, with a very conservative clockspeed (to showcase their performance/W metric), which I would expect they could ramp up for more performance on the cost of worse performance/W (nvidias are already showned to run at incredible high clockspeed from the beginning). The biggest unknown metrics right know is the advancement in GCN 1.3 (and what influence it will have on performance) and the final clockspeed on the individual SKUs.

 

Now, we are all just waiting patiently for nvidia to spoil is with some actual numbers (not theorized numbers they pulled so far), and to actual showcase silicon running software.

They have been awfully quite from that camp.

[79wjd]

2 hours ago, Citadelen said:

Nvidia didn't have the expertise to take up the contract, nor did they have the manufacturing capacity.

based on..... 

[Tomsen]

7 minutes ago, djdwosk97 said:

based on..... 

Nvidia doesn't offer any kind of semi-custom x86 SoC solution, and doesn't offer any kind of custom SoC solution.

IE Nvidia didn't have any comparable solution for the console manufactures.

[79wjd]

4 minutes ago, Tomsen said:

Nvidia doesn't offer any kind of semi-custom x86 SoC solution, and doesn't offer any kind of custom SoC solution.

IE Nvidia didn't have any comparable solution for the console manufactures.

They never have, yet they got the PS3 contract. The point is, the consoles easily could have gone with a different SOC but chose not to due to cost reasons (hence why the consoles aren't particularly profitable for AMD).

[Tomsen]

1 minute ago, djdwosk97 said:

They never have, yet they got the PS3 contract. The point is, the consoles easily could have gone with a different SOC but chose not to due to cost reasons (hence why the consoles aren't particularly profitable for AMD).

The PS3 used a dGPU, not an integrated one. A huge point was that console manufactures wanted a real SoC, with almost-all/all components integrated. Something Nvidia can't offer today. Another point would be x86, which Nvidia can't offer. Sure, they could continue with external processors, but that would not be because of cost reasons (as external processors are MORE expensive).

AMD consoles have been very profitable for AMD, without it, they most likely wouldn't be here today. What makes you think otherwise?

[Nineshadow]

17 hours ago, Aytex said:

 

If this is true i can only imagine the power of the flagships, maybe like 3x the power? lol if we can only imagine

it also consuming less TDP makes us not have to get upgraded PSU's= more money saved for us

Source

It's price/performance.

At $300, it has twice the price/performance of the 980Ti which is priced at $600. This means that it must perform about the same as a 980Ti.

 

Although I doubt the 490X will cost just $300.

[Citadelen]

24 minutes ago, djdwosk97 said:

based on..... 

AMD has access to both TSMC and Global Foundrie, and Nvidia has mediocre CPUs that are ARM, and not x86 and AMD has expertise in APUs which suited Microsoft and Sony perfectly.

[Commander Llama]

That's not even a very nice looking made up graph...

[Tam3n]

Good, but the vertical axis of the chart is named upside down.

[Phoenix721]

1. The Polaris 10 is the flagship, isn't it?
2. Good luck vs the fury x in 4k with the GDDR5x or whatever it is called memory.

[Valentyn]

Just now, Phoenix721 said:

1. The Polaris 10 is the flagship, isn't it?
2. Good luck vs the fury x in 4k with the GDDR5x or whatever it is called memory.

Vega is the flagship, Polaris 10 is Mainstream, and P11 is entrance/mobile

[Phoenix721]

Just now, Valentyn said:

Vega is the flagship, Polaris 10 is Mainstream, and P11 is entrance/mobile

Vega is not coming out untill 2017 with a different architecture. Also, vega will have HBM2 and not freaking gddr5x.

[Notional]

Just now, Phoenix721 said:

1. The Polaris 10 is the flagship, isn't it?
2. Good luck vs the fury x in 4k with the GDDR5x or whatever it is called memory.

Vega is the flag ship with HBM 2.

 

Polaris 10 is apparently larger than 11, so that would make it the middle chip (similar to 970/980 model wise).

[79wjd]

7 minutes ago, Phoenix721 said:

2. Good luck vs the fury x in 4k with the GDDR5x or whatever it is called memory.

But you don't need HBM for 4k

[patrickjp93]

20 hours ago, Agost said:

I'm actually quite hoping so.

They listed the full GP100 at 300W, highest ever for a NVIDIA GPU, while it has ~20% more cores than GM100 and the same area...

And 40% higher clocks*

 

1400MHz boost clock is massive.

[Phoenix721]

1 minute ago, djdwosk97 said:

But you don't need HBM for 4k

you don't, but in AMD's case, it really, really really helps.

[patrickjp93]

1 minute ago, djdwosk97 said:

But you don't need HBM for 4k

Shh... the plebes still don't understand it's all about the cache anyway. It may come to matter eventually, but games are not so complex currently (nor will be for a few years to come) that they actually require 512GB/s of bandwidth from VRAM, much less 1 TB/s.