Gemini (FuryX2) looms near

[railven]

That means you think AMD will bring down Fury X level of performance from $649 to $299-349 in 12 months? Those days seem to be long gone. Welcome to the milking way of selling graphics cards => 25-35% performance increases trickled down slowly while milking each card as a flagship at $550-650.

No, more so I was expecting the mid-range cards to be 15-20% faster, ie a 390X + 10-15% due to uarch tweaks and node change. Which would put it at about 80% of a Fury X @ 4K, and well into Fury X territory @ 1080P.

Two of those little guys combined might have some unused power, say even 10-15% OC headroom. Because something tells me Fury X2 is going to be closer to 2xNanos than 2xFury Xs and suddenly you got an option comparable to Fury X2. But without any concrete numbers, just speculation .

The way both companies set up a GPU generation now makes it almost impossible to "win." On the front end of a generation, you'll get milked for mid-range die at flagship prices, on the back end of a generation, you'll get milked for flagship die at flagship prices but then it's already almost half-way into the generation. I offer my 6 ways to upgrade with the new system:

Your 6 steps of upgrading probably explains why I'm so happy with my purchases. Hell, 5/6 things you list I've been doing for years.

I'm already itching for a new upgrade ANd the funds have already been allocated. Plus, the GF's 780 Lightning is going to go to my nephews while the GF gets the nice water cooled heavily OC'ed 980 Ti.

Upgrades for everyone!!!

 

[maddie]

Profits matter more than market share long-term. It's better to be profitable with 18% market share than to have 50% market share by essentially giving away your product.

Also, as I already stated, showing up first is more important for AMD. Let's say there are 1000 PC gamers, of which 300 are 100% upgrading their GPU in 2016. Of these, 40% are loyal to team green, 50% brand agnostic, 10% loyal to team red. If AMD launches late, they will lose 90% of 300 customers. If you launch early, you can go for 50% of brand agnostic. Rory Read is the one who knew this about the GPU market. Objective/brand agnostic PC gamers won't wait 6-9 months for AMD's response and OEMs won't either. That's why for AMD, winning back market share has to be prioritizing launch timing, more than absolute max performance.

Profits are important/vital but it has to be not only % profit but the actual quantum in order to continue funding R&D, etc.

18% market-share with Nvidia is a way to die slowly. They must have good % profits AND increased market-share to offer future competitive products.

Agreed, an early launch is very important.

 

[moonbogg]

The ones that waited are the AMD loyalists, and objective types who were not in any hurry to make a choice. Everyone else just grabbed a 980ti.

If you were smart, you grabbed two incase of a future shortage.

 

[MrTeal]

No, more so I was expecting the mid-range cards to be 15-20% faster, ie a 390X + 10-15% due to uarch tweaks and node change. Which would put it at about 80% of a Fury X @ 4K, and well into Fury X territory @ 1080P.

Two of those little guys combined might have some unused power, say even 10-15% OC headroom. Because something tells me Fury X2 is going to be closer to 2xNanos than 2xFury Xs and suddenly you got an option comparable to Fury X2. But without any concrete numbers, just speculation .

Ah, it was a little unclear what you were referring to. Node shrink + uarch wouldn't bring a 100-120mm² over Hawaii levels of performance. Even if the demo was capped and the chip has twice as much power on tap, they're still comparing it with a 950 which is R9 270X levels of performance. Hawaii is over twice as large as Pitcairn, that's a pretty massive gap to make up.
Two midrange chips in the 200mm² range is another story though, I could see those beating Hawaii and nipping at the tails of Fiji.

Really, if this is a 100-120mm² die, it raises some interesting questions about AMD's strategy. GCN1 had Cape Verde @ 123mm², Pitcairn @ 212mm², and Tahiti @ 352mm². Launching a small die makes a lot of sense, especially to push into mobile and combat the existing GM107 laptops if they can launch first. If they are only launching one other die in 2016 though, what is it going to look like? I always figured we'd see a top die in the 300-350mm² range, but that leaves a massive gap in their lineup. If they launch something Pitcairn size, they likely aren't even going to be able to beat Fiji in current games, which would be unheard of flagship to flagship for such a large node jump.

I can't wait to find out more information on what the Polaris stack will look like. A lot of the speculation is based on Ryan's estimate of die size, so the margin for error could be pretty huge. 100-120mm² would by my pinky nail size, my middle finger nail would be 190-200mm². It wouldn't be tough to be out by a decent margin and make all our idle guesswork for naught.

 

[maddie]

Ah, it was a little unclear what you were referring to. Node shrink + uarch wouldn't bring a 100-120mm² over Hawaii levels of performance. Even if the demo was capped and the chip has twice as much power on tap, they're still comparing it with a 950 which is R9 270X levels of performance. Hawaii is over twice as large as Pitcairn, that's a pretty massive gap to make up.
Two midrange chips in the 200mm² range is another story though, I could see those beating Hawaii and nipping at the tails of Fiji.

Really, if this is a 100-120mm² die, it raises some interesting questions about AMD's strategy. GCN1 had Cape Verde @ 123mm², Pitcairn @ 212mm², and Tahiti @ 352mm². Launching a small die makes a lot of sense, especially to push into mobile and combat the existing GM107 laptops if they can launch first. If they are only launching one other die in 2016 though, what is it going to look like? I always figured we'd see a top die in the 300-350mm² range, but that leaves a massive gap in their lineup. If they launch something Pitcairn size, they likely aren't even going to be able to beat Fiji in current games, which would be unheard of flagship to flagship for such a large node jump.

I can't wait to find out more information on what the Polaris stack will look like. A lot of the speculation is based on Ryan's estimate of die size, so the margin for error could be pretty huge. 100-120mm² would by my pinky nail size, my middle finger nail would be 190-200mm². It wouldn't be tough to be out by a decent margin and make all our idle guesswork for naught.

Black bold:
We are missing something important. What?

Red bold:
According to Pcper, this small die is expected to sell into the GTX 950 market range, so I would think it won't be be a 200mm^2 14nm die. Also probably too big for the thin & light notebook market. A guess of course.

 

[3DVagabond]

I disagree. Plenty of people, a lot even on these forums, waited for Fury X to launch, even though the 980 Ti had already been released. I know this goes against your "buyers are sheep" narrative though. As far as the 6-9 month waiting period, who knows if people will wait that long. I don't see there being that large of a gap between releases...

He didn't say nobody would wait. Of course some will. That's not enough for AMD though.

 

[Mondozei]

gamers will upgrade 2 or even 3 times in the same generation

You still treat this as an instance of corporate greed. You should know better. From our own Ryan Smith:

the regular march of progress in semiconductor fabrication has quickly tapered off over the last decade. What was once a yearly cadence of new manufacturing processes – a major new node every 2 years with a smaller step in the intermediate years – became just every two years. And even then, after the 20nm planar process proved unsuitable for GPUs due to leakage, we are now to our fifth year of 28nm planar as the leading manufacturing node for GPUs. The failure of 20nm has essentially stalled GPU manufacturing improvements, and in RTG’s case resulted in GPUs being canceled and features delayed to accommodate the unexpected stall at 28nm.

So to say that the forthcoming move to FinFET for new GPUs is a welcome change is an understatement; after nearly half a decade of 28nm GPUs we finally will see the kind of true generational improvements that can only come from a new manufacturing node.

So we've had half a decade on the same node, so why are you still feigning surprise that people upgrade 2 times per generation? That comes out as once every 2.5 years for the average person.

And as Ryan noted, the rapid advances in node shrinks meant you could sell the high-end GPU right away and focus on the next the year after, because you either had a full node shrink or a half-shrink.

Now we don't get one for 5 years.

RS, it's tiring to read your ignorance, especially when you often write decently intelligent stuff on most issues. Both AMD/NV have to adapt to the new reality of a dead Moore's law. You still show that you haven't.

 

[96Firebird]

He didn't say nobody would wait. Of course some will. That's not enough for AMD though.

He said 90% wouldn't wait for AMD. I think that number is way too high.

 

[exar333]

I probably would have purchased the X2 if it released last fall. It is arriving too late honestly. As others have said, most people just already bought a 980Ti or Fury X. With 14/16nm cards just around the corner, spending $1000+ on this seems really premature. Not saying it will not be a great card, but definitely worth waiting and seeing in regards to Pascal and Polaris.

 

[maddie]

I probably would have purchased the X2 if it released last fall. It is arriving too late honestly. As others have said, most people just already bought a 980Ti or Fury X. With 14/16nm cards just around the corner, spending $1000+ on this seems really premature. Not saying it will not be a great card, but definitely worth waiting and seeing in regards to Pascal and Polaris.

What is the entire world market for SLI/Xfire? I remember reading 10's of thousands somewhere. Not a big market segment. This is more a bragging right/public image market segment, not a money machine.

What I want to see is the initial reviews using many video cards with the Rift. This is certainly what AMD is looking to use as a booster for their public image. They would have tested Occulus VR with all of the top end cards by now and by delaying release to coincide with VR shipments, appear to feel confident of good reviews.

 

[tential]

He said 90% wouldn't wait for AMD. I think that number is way too high.

Ya but still agree the amount of people that will wait will not be enough for amd.

 

[n0x1ous]

What I want to see is the initial reviews using many video cards with the Rift

This is going to be the driver for multi-gpu moreso than any of the DX12 low level non AFR stuff. VR + multi gpu are match made in heaven.

 

[3DVagabond]

He said 90% wouldn't wait for AMD. I think that number is way too high.

What's your calculation based on? You simply disagree with no basis. You seem to be simply muddying the waters.

 

[96Firebird]

What's your calculation based on? You simply disagree with no basis. You seem to be simply muddying the waters.

If AMD launches late, they will lose 90% of 300 customers.

You may have missed that? He explicitly says 90%.

I disagree that 90% wouldn't wait for AMD. I am not going to just make up numbers out of thin air like he did to supply you with a "calculation". Why didn't you ask for the basis of his numbers?