[Rumor (Various)] AMD R7/9 3xx / Fiji / Fury

[monstercameron]

I suspected that as well, but how does this explain the fact that the R9 M295X (full Tonga) is also being sold to other vendors? Dell uses it in some Alienware laptops.



Agreed. Given AMD's R&D limits, they would have been better off putting the money into two smaller chips instead of Fiji. Especially since Fiji uses HBM 1st generation technology, which not only limits it to 4GB (a real problem for a product positioned at the ultra high end), but also will be obsolete next year when HBM2 comes along. Those defending AMD's decision argue that it would be a waste to invest in a whole new 28nm lineup when it's going to be superseded soon, but that logic applies much more strongly to investing in HBM1, and that investment probably cost AMD much more money - money that it clearly can't afford to waste.

The only things amd are lacking seems to be a halo product and better power efficiency from a marketing standpoint. All their products are still very competitive in terms of performance.

If the choice was redesign 10 different chips on an aging node vs refreshing the mainstream and beta testing new memory spec, one seems less risky than the other.
Also when did 4gb become inadequate for 4k gaming?

 

[ShintaiDK]

The negativity on the rebrands is understandable but declaring Fiji a failure before it ever comes out is silly at best.

Economicly Fiji is a failure in the current situation. Thats undisputeable when we see the rest of the lineup that is actually supposed to ship in the high volume and revenue unlike the Fury cards.

 

[ShintaiDK]

Agreed. Given AMD's R&D limits, they would have been better off putting the money into two smaller chips instead of Fiji. Especially since Fiji uses HBM 1st generation technology, which not only limits it to 4GB (a real problem for a product positioned at the ultra high end), but also will be obsolete next year when HBM2 comes along. Those defending AMD's decision argue that it would be a waste to invest in a whole new 28nm lineup when it's going to be superseded soon, but that logic applies much more strongly to investing in HBM1, and that investment probably cost AMD much more money - money that it clearly can't afford to waste.

We saw a bit of the same with Phenom. "Native quadcore" was much more important than common sense. More important to be first rather than something you can make money on. HBM1 is turning into a nightmare for AMD. And they only make it worse themselves with the 8GB 390/390X offerings.

 

[3DVagabond]

Between OpenCL and all of the rendering API's being Mantle based, I'd say that nVidia has to be a bit worried about their little closed vendor empire.

http://videocardz.com/55081/nvidia-geforce-gtx-titan-x-opencl-performance-leaks-out

http://www.phoronix.com/scan.php?page=article&item=nvidia-amd-opencl-2015&num=2

https://compubench.com/result.jsp?b...=true&arch-ARM=true&arch-x86=true&base=device

I don't know man. Nvidia and AMD both seem to be doing well in OpenCL benches. You can perhaps find some left field OCL bench that skews the whole test here and there for either side, but I don't think Nvidia has anything to worry about.

nVidia's relative performance has nothing to do with my statement. In your rush to defend you read into what I said what you wanted and didn't actually look at what I said. Remember this the next time I try and explain to you that you are claiming I'm saying something that I am not. You do this quite a bit.

The fact is nVidia likes to control and manipulate everything. Lock people out and make sure they have an advantage. That's what has made nVidia what it is today. You can't do that with open source. Without them being able to stack the deck at the software level they will have to actually compete solely on the hardware level.

Besides your statement about "You can perhaps find some left field OCL bench that skews the whole test" is accusational and unnecessary. I am not doing anything like that. Especially when you are the one who retorts with benchmarks of your choosing. We'll see how much nVidia has caught up when Fury comes out and we get full OpenCL 2.0 benchmark suite results and, more importantly, tests in real world programs that people actually use.

 

[boozzer]

16th can't get here fast enough 🙂

 

[EightySix Four]

Economicly Fiji is a failure in the current situation. Thats undisputeable when we see the rest of the lineup that is actually supposed to ship in the high volume and revenue unlike the Fury cards.

How could we possibly know if it is a failure? If it's faster than a 970 (not hard to do), actually has 4gb+ of vram, and comes in at the same price or lower, it's no question a win even if it's a rebrand. Economically all that matter is performance/$ and how much it costs AMD, and you don't know either of those pieces of information so...

 

[ShintaiDK]

How could we possibly know if it is a failure? If it's faster than a 970 (not hard to do), actually has 4gb+ of vram, and comes in at the same price or lower, it's no question a win even if it's a rebrand. Economically all that matter is performance/$ and how much it costs AMD, and you don't know either of those pieces of information so...

Fury is a new design, not a rebrand. But it will sell miniscule compared to other chips. Specially the 370 and down. Chips(250/270) that desperately needed to be moved away from GCN 1.0.

 

[EightySix Four]

Fury is a new design, not a rebrand. But it will sell miniscule compared to other chips. Specially the 370 and down. Chips(250/270) that desperately needed to be moved away from GCN 1.0.

Okay, then I'm confused. Are you saying Fiji is an economic failure like you said here:

Economicly Fiji is a failure in the current situation

Since that simply doesn't make sense as we don't know the economics of the chip (performance/$, cost for AMD to make)...

Or are you saying the 3** series is a failure because they're rebrands which you are trying to say here:

But it will sell miniscule compared to other chips. Specially the 370 and down. Chips(250/270) that desperately needed to be moved away from GCN 1.0

Which doesn't make sense since we don't know what their performance will be compared to nVidia's lineup at each pricepoint...

 

[3DVagabond]

Sapphire has 390X, 390 and 380 on their website. There is also section for R9 FuryX and R9 Fury, but those are still blank pages.

Sapphire Tri-X R9 390X 8G D5; https://www1.sapphiretech.com/productdetial.asp?pid=D40475DB-8BD0-40F6-8C33-F12D63272AEC&lang=eng

Edit: "Power Consumption: 375W" according Sapphire for 390X :O

390X

390

Interesting they've changed the outputs. The 3x DP might indicate being able to run 3x4K surround. Also might possibly mean HDMI 2.0? Although looking at the specs there's nothing to indicate something one way or the other.

 

[ShintaiDK]

Okay, then I'm confused. Are you saying Fiji is an economic failure like you said here:



Or are you saying the 3** series is a failure because they're rebrands which you are trying to say here:

Well both got problems. But if you got a limited amount of money like AMD. Spending them on Fury rather than updating 250 and 270 is bad business. Because Fury is that part that will make the least amount of money for AMD.

 

[.vodka]

We know a 290x can boot with the 390x bios that got captured a few pages back... so there's not much difference between both cards and boards.

If anything any new features will be mostly software based, and should be able to be used on the 200 series too if logic applies.

 

[maddie]

we'll know who doesn't know what they're talking about if they try to use this 375W figure in their arguments

:thumbsup:


Didn't take long.😛

Set up. Knocked down.

Now I feel okay moving Shintai to ignore.


As for these cards, I await reviews, but I was in the market for Fury or X vs 980ti anyways.

This thread is a blast.

To paraphrase Sir Winston Churchill.

'Never in the history of human debate was so many disgusted by so few.'

 

[EightySix Four]

Well both got problems. But if you got a limited amount of money like AMD. Spending them on Fury rather than updating 250 and 270 is bad business. Because Fury is that part that will make the least amount of money for AMD.

No no no, see, you can't just focus on low end development, that's when you get your ass kicked the next generation. You have to focus on building the next generation of tech that will trickle down. If AMD thinks they can drag another year out of their current range (while remaining competitive perf/$) while developing new tech they will use at 14nm, then that's the logical thing to do over a multi-year period.

 

[Skurge]

Okay, then I'm confused. Are you saying Fiji is an economic failure like you said here:



Since that simply doesn't make sense as we don't know the economics of the chip (performance/$, cost for AMD to make)...

Or are you saying the 3** series is a failure because they're rebrands which you are trying to say here:



Which doesn't make sense since we don't know what their performance will be compared to nVidia's lineup at each pricepoint...

I think he meant that the idea of only updating the halo product was a bad idea.

In any case. the large turning points for both vendors was their halo cards. 4600ti, 9700 Pro, 7800gtx, 8800gtx, 4870, gtx980 and their failures even when their cheaper cards were good brought down the whole generation. The point of a halo product is for the cheaper stuff to piggyback on the impression the halo cards have brought.

So if Fiji turns out to be the best thing since sliced bread, the perception of AMD cards as a whole will be improved, rebrands or not. If that is enough to recapture the lost marketshare... I don't think any of us armchair experts will know.

 

[ShintaiDK]

No no no, see, you can't just focus on low end development, that's when you get your ass kicked the next generation. You have to focus on building the next generation of tech that will trickle down. If AMD thinks they can drag another year out of their current range (while remaining competitive perf/$) while developing new tech they will use at 14nm, then that's the logical thing to do over a multi-year period.

Fury isnt a new tech besides HBM. And tech havent trickled down. The 250 and 270 aka 370 is a prime example of this. And this is one of the reasons AMD is bleeding marketshare at an alarming rate. 35% to 22.5% in 1 year.

Tonga introduced GCN 1.2. The only GCN capable of 4K video decode as well. Almost a year later and what products got it besides Tonga? Unreleased Fury and unreleased Carrizo.

And dont expect any new GPU tech at 14nm from AMD. It looks to be GCN 1.2 shrinks at best when listening to their financial analyst day.

 

[AtenRa]

No no no, see, you can't just focus on low end development, that's when you get your ass kicked the next generation. You have to focus on building the next generation of tech that will trickle down. If AMD thinks they can drag another year out of their current range (while remaining competitive perf/$) while developing new tech they will use at 14nm, then that's the logical thing to do over a multi-year period.

Dont try to reason with him, if AMD would spend money for the low end cards, he would say AMD should of spend money on the high end where the big profits are. 😛

 

[.vodka]

Fury isnt a new tech besides HBM. And tech havent trickled down. The 250 and 270 aka 370 is a prime example of this. And this is one of the reasons AMD is bleeding marketshare at an alarming rate. 35% to 22.5% in 1 year.

You don't know that. They could've easily gone further than Tonga, improving GCN even more.

We'll see in two days.

 

[EightySix Four]

Fury isnt a new tech besides HBM. And tech havent trickled down. The 250 and 270 aka 370 is a prime example of this. And this is one of the reasons AMD is bleeding marketshare at an alarming rate. 35% to 22.5% in 1 year.

Tonga introduced GCN 1.2. The only GCN capable of 4K video decode as well. Almost a year later and what products got it? Unreleased Fury and unreleased Carrizo.

First, we don't know if there are any GCN improvements in this generation or not yet, so you're basing this entirely off speculation and rumor. Second, that's not what's going to sell cards right now, what's going to sell cards if launching a range where for every $ point, AMDs cards outperform their nVidia counterparts. Especially if they have a halo card that brings back some attention.

AMD went from 35% to 22.5% in one generation because they failed to do that initially, and then once they caught up the original benchmarks and reviews still hurt them too much. A 290x is the best value on the market right now, period. Relaunching it with a new name, a slightly lower TDP, and a similar price would be considered a success by everyone who doesn't read these forums everyday.

 

[3DVagabond]

I guess this confirms 2 Fury cards. Looking foward too se the performance of the Fury X and Fury.

There are reports that the WCE version is AMD only, not the AIB's. So, there might be 3 versions.

 

[AtenRa]

I think he meant that the idea of only updating the halo product was a bad idea.

In any case. the large turning points for both vendors was their halo cards. 4600ti, 9700 Pro, 7800gtx, 8800gtx, 4870, gtx980 and their failures even when their cheaper cards were good brought down the whole generation. The point of a halo product is for the cheaper stuff to piggyback on the impression the halo cards have brought.

So if Fiji turns out to be the best thing since sliced bread, the perception of AMD cards as a whole will be improved, rebrands or not. If that is enough to recapture the lost marketshare... I don't think any of us armchair experts will know.

I dont believe they give a damn about market share, they prefer to have 20-30% market share and make a profit than have %40-50 market share and bleed money.
NVIDIA have shown High End gamers are willing to spend $500 to $1000 or more (SLI/CF). A high end GPU also gives you a Professional GPGPU card with even higher margins. Why spend money on the low end to gain market share when you really need to make a profit first ??

 

[Skurge]

No no no, see, you can't just focus on low end development, that's when you get your ass kicked the next generation. You have to focus on building the next generation of tech that will trickle down. If AMD thinks they can drag another year out of their current range (while remaining competitive perf/$) while developing new tech they will use at 14nm, then that's the logical thing to do over a multi-year period.

Developing for the low and midrange is what got AMD into this mess anyway. Since 2900XT AMD has been focusing on the "sweet spot" and the "midrange" cards. Focusing on the $199 and $299 range. When they were focusing on the highend 9700 pro, X800, X1900, it didn't matter how good or bad their midrange stuff was. People still bought it, cause it came from the same guys that built the fastest GPUs, not the cheapest.

Remember this? a lot of good that did them.

 

[TaintedSquirrel]

There are more 300-series leaks... Not even worth the trouble posting anymore, who cares.
Bring on Fury.

 

[happy medium]

Guys has one (390x) over at Hard forum, benching and snapping pictures.

http://hardforum.com/showthread.php?t=1865280&page=2

Here is a link to the 390x bios dump if you guys wanna flash your 8gb 290x's. 🙂

http://s000.tinyupload.com/index.php?file_id=85146925914006580479

No HDMI 2.0 support

 

[flopper]

I.

I wouldn't take that 375W figure too seriously.

You figure people know about marketing by now to not say what the max draw is by specs but actual use, I know its not 375w BUT that wont matter as all I see is 375w and that will soon be stated to be with the 390 series even worse than last time.
do the same as nvidia state a different tdp how hard can it be?
its writing a few different numbers ffs.

they need to fire a few at amd camp and hire someone who knows more about perceptions.

 

[Azix]

Economicly Fiji is a failure in the current situation. Thats undisputeable when we see the rest of the lineup that is actually supposed to ship in the high volume and revenue unlike the Fury cards.

We saw a bit of the same with Phenom. "Native quadcore" was much more important than common sense. More important to be first rather than something you can make money on. HBM1 is turning into a nightmare for AMD. And they only make it worse themselves with the 8GB 390/390X offerings.

where on earth are these assertions coming from? 😵 😕 😵

I doubt people are taking you seriously knowing your apocalyptic leanings when it comes to AMD and the GPU market.

Fury pro will sell well if under $500. Better if significantly under 500. the e3 conference has a chance of turning things around on the marketing front as well. AMD could make themselves the manufacturer to have for dx 12 and the future of PC gaming.