The charts also have the XFX model which operates at reference clocks, which I agree is the better comparison. Sorry if that was unclear.
@NTMBK
Your comparing a aftermarket ASUS DCUII OC 280x to a stock clock reference GTX 770 in those BM charts, not a good/fair comparison imo.
Bionare vs maxwell? So you pick the only two recent architectures that could possibly make nvidia look decent and want me to take that comparison as a prediction on 20nm? Ridiculous!
It's pretty simple… GK110 and hawaii are the respective high end 28nm parts, no? How is it even close to relevant to compare bionare to nvidias current newest chip? GK110 = 560mm. Hawaii = 435 mm. GK110 5.2 GFLOPS SP Hawaii = 5.7 GFLOPS SP
Hawaii = .0131 GFLOPs/mm^2
GK110 = .0093 GFLOPs/mm^2
If hawaii was 560mm^2 it would be 7.3 GFLOPs. That is how much better AMD is at compute, and I won't even mention, DP it gets ugly for nvidia there.
AMD purposely made hawaii a compute card, look at the W9100. If they want, they can grab the crown from nvidia any time they want.
As far as "bias", my only bias comes from owning both products for years. I think nvidia makes good products sometimes (6800,8800,285) but they always overcharge and they aren't as good with hardware as AMD. Do you own any AMD products right now?
The charts also have the XFX model which operates at reference clocks, which I agree is the better comparison. Sorry if that was unclear.
you realy cannot talk about performance and die size without transistors and density. Die size really has little to do with performance at all. Its important for financial reasons, I guess. But your trying to give it some special value or make it a winning metric. I mean, I am sure if we want we can find all sorts of ways to do this. Like ounces or what ever.
Example, "AMD 290x weighs 3 onces less than GTX 780ti so AMD wins in the per ounce metric. If AMD cards wieghed as much as the 780ti they would be x times more powerful."
Why not?
yeah, right.
Die size means very little because it is the transistors that do the work. AMD crams as many transistors into a given area in order to save money. This allows them to sell their chips cheaper. Your scenario of a chip the size of the GK110 is impossible, completely impossible. AMD is at the thermal and power limits already with their 435mm2 chip. So ask yourself, how is it that AMD reached the limit at 431mm^2 and Nvidia made it all the way up to 560? Bingo!!!
But lets talk about transistors.
Its not only the amount crammed into the area (described as density) but also the type and layout. This is where they are completely not even comparable. Its what makes these architectures radically different. The Hawaii has 6.2 billion vs the gk110s 7.1billion.
Hawaii 14million transistors per MM^2
Gk110 12million
AMDs design is packed in more dense. Its an achievement that saves them money. That is great. But.........
Its not that simple. This is still rather meaningless.
See, you cannot just magically turn Hawaii into a 560mm^2 chip. That is insane. That is just not how it works. Even if it did, there is a huge power constraint preventing this. Chips are designed and engineered painstakingly. The choices dictated the outcome. The end product is a result of many years worth of work.
Hawaii is a great achievement but it is drastically different than the gk110. Different paths, purposes, and end with different products.
They can pack more transistors and drop clocks, that way they may still win even more performance and regain some perf/watt.
AMD probably designed Hawaii tuning transistor for area density, meanwhile Nvidia did it for performance. IMO AMD's road paid out as they could achieve more density and a smaller die with 95% of the performance of GK110 without compromising the thermals and power consumption that much. At least, that is for gaming. For professional workloads, its a engineering marvel to have full FP64 rate in such a crammed design.
Now, die size or transistor density can indirectly affect things. Heat. Efficiency. Those two affect noise output, depending on cooling solution. I could go on here. That can affect the final product, and that's what people buy based on. It's not like "oh BABY this is only 420mm^2, i'm buying that on day one!" Perhaps i'm the only one that finds this entire die size thing a bit silly. I just look at the end results and features that I care about. I'd say most consumers are the same. GM200 can be 200mm^2 or 900mm^2 for all I care. If it performs like a beast with the same nvidia intangibles which Kepler has had, I want one.
Usually with a new node poor yields do happen.
Plus if AMD is going to use stacked DRAM and all that fancy stuff, it might take some time.
Considering Hawaii came out end of October/ beginning if November 2013, Q2 2015 with a 1.5 year delay seems plausible to me.
Just out of curiosity, do you have a good solid source on this? Because as far as I can tell, GPUs are in the planning phase for more than a year prior to launch. I don't think 20nm was in consideration given that timeframe, but I could be wrong. In any case, if you have a link to a good solid source to shed light on that.
I do know that cayman was affected by something similar, because 32nm was basically dropped by TSMC. But I haven't heard anything about Hawaii being planned for 20nm from the outset - as it was known for some time that 20nm would not be ready at any point in 2013. And that would definitely overlap with the planning and engineering phase for the GPU. My thoughts are that AMD knew from the get-go that 20nm would be an absolute no go. But if i'm wrong, a *good* source (ie, not semiaccurate or other national enquirer type tech news sites) would be great, i'm really curious about this.
So, we ignore the simple fact that AMD can't hold their 1000MHz and need to throttle it down to <850MHz? :hmm:
Unfortunately, I do not have a source that directly says that. I know they've delayed their 20nm significantly, so its likely when AMD's engineers started the design phase of Hawaii they thought 20nm would be available. When it was delayed partway through Hawaii's design, they had to make do with 28nm.
Well, the better density of Hawaii is a decent part of the reasons the 290 and 290X performance per $ of MSRP was good compared to Nvidia's high end. Performance/$ is important to a significant number of potential GPU buyers. Also given the focus of TSMC's 20nm on power efficiency (not pushing clock speeds), density will be one of the main ways to get more performance for a given die size. Given this thread is about AMD 20nm, I'd say the Tahiti->Hawaii same node density boost is something worth talking about.
I'm trying to think of the last time I bought a GPU based on die size.
..........still thinking.
Nope.
You guys can argue this all day long but people buy based on software, features, performance, and value adds. To that end I prefer nvidia and i'm not shy about stating such, but if someone prefers AMD that's cool too. That's what competition is all about. Clearly AMD is making more of an effort to do value adds with their GPUs, whether they're successful in that .. time will tell. I'm just not sure where die size fits into the purchasing equation. I mean does a buyer get some sort of psychological rush from knowing they have a smaller die on their GPU. I doubt it.
Now, die size or transistor density can indirectly affect things. Heat. Efficiency. Those two affect noise output, depending on cooling solution. I could go on here. That can affect the final product, and that's what people buy based on. It's not like "oh BABY this is only 420mm^2, i'm buying that on day one!" Perhaps i'm the only one that finds this entire die size thing a bit silly. I just look at the end results and features that I care about. I'd say most consumers are the same. GM200 can be 200mm^2 or 900mm^2 for all I care. If it performs like a beast with the same nvidia intangibles which Kepler has had, I want one.
I'm not so sure. I'm sure it's good for AMD, but is it good for the buyer? It's a question worth asking perhaps. Maybe density also affects heat output and efficiency, so you have good price/performance at the baseline but by the time the final product rolls around, it's hot and loud so you're relegated to buying an aftermarket card at a higher price point. By then, the price performance picture isn't so clear since the aftermarket variety might cost 70-80$ more than a reference design.
So it aint quite that simple. If the density negatively affects efficiency and heat output, then it is one step forward and two steps back. And that more robust cooler offsets any potential price/performance in the first place. I'm just saying.
I'm not so sure. Maybe density also affects heat output and efficiency, so you have good price/performance at the baseline but by the time the final product rolls around, it's hot and loud so you're relegated to buying an aftermarket card at a higher price point. By then, the price performance picture isn't so clear since the aftermarket variety might cost 70-80$ more than a reference design.
So it aint quite that simple. If the density negatively affects efficiency and heat output, then it is one step forward and two steps back. And that more robust cooler offsets any potential price/performance in the first place. I'm just saying. I don't think it's a black and white issue. Maybe it can be better, maybe not. Maybe more dense isn't always better, maybe bigger die size or smaller die size isn't always better. It just depends really. I think AMD did what they did for increased profitability. Which is fine really, that's what corporations do. They get more dies per wafer. Good on them. But is it good for a consumer? Well, again, not a black and white issue. If more dense == more heat, then clearly it doesn't benefit the consumer. If more dense == less efficiency, that means AIB makers will spend more on their custom solutions. So the picture becomes not so clear at that point. Regardless, the bottom line is nobody makes a buying decision on die size.
You mean up to 10 fps? First of all, as you increase the resolution the performance graphs "compress" along with that increase in resolution. So in fact, yes, it is a commanding lead; many titles are 5-10 fps lead in favor of the 780ti compared to an aftermarket 290X (see the legitreviews link I posted earlier. They tested the Powercolor PCS triple slot card, which so far seems to be one of the fastest/best custom 290X cards).
So performance graphs get "compressed" as the resolution goes up. As an example, let's say you do benchmarks at 1080p. On average, maybe the 780ti has a 20-25 fps lead over the competition at 1080p. When you increase the resolution, the performance graphs compress a tad. At 2560x1600, the 20-25 1080p performance lead becomes 15 fps or so. At 4k? The lead becomes 5-10 fps. Yes, at 4k that is a commanding performance lead. And yes, while I don't think there's quiet mode on the Powercolor PCS custom 290X, I know it did not throttle or anything like that before you want to theorize such. It's an aftermarket custom card. View the benchmarks at 4k for yourself at legitreviews. Their powercolor PCS review is on the front page. No, it didn't throttle. No, it didn't use a clockspeed gimped quiet mode - to answer all of the questions brought up earlier.
That's not to say that the 290X is a slouch, it's clearly a very fast GPU and is cheaper. The 290X is a fast GPU in its own right and has more of the value and bang for the buck angle which AMD is generally known for. You likely won't be disappointed with either GPU at 4k resolution. My main point was the MST driver bug with NV has been resolved and has given Nv a significant 4k resolution performance bump. Obviously, a GPU being gimped at 70% GPU utilization will cause it to lose benchmarks - that was the old NV driver bug with MST, which has been fixed.
In any case, this is somewhat of a tangent meant to answer some of the questions in this thread. So with those questions answered, i'll leave that off topic alone, done talking about it. Anyway, about 20nm. If I may, the state of 4k resolution performance clearly indicates that 20nm may be the answer for that. So i'm pretty excited and optimistic for 20nm GPUs and Maxwell - they should be that much better for 4k resolution performance, perhaps at 20nm we can use a single GPU without having to compromise settings to get 60 fps in most titles. Of course you can do that now if you fiddle with settings a bit, but there are some out there who like doing ultra settings across the board. Maybe 20nm can bring us that. At 4k resolution. Without crossfire or SLI. One can hope. In any case, it will be interesting to see what happens with 20nm and next gen GPUs later this year.
I see maybe two games in all those reviews where it was greater than 5 fps different and that was arkham origins which by all reports was intentionally gimped on amd cards by the developer. They also don't use the highest available settings in that test which I believe the 290x would have trucked through with its 512 bus.
Guru3d showed your commanding lead to be 0-5 fps ( at a glance) in all games with two games using reduced IQ settings where 780ti has a greater than 5 fps lead.
Better yet, buy two 290x and two 780tis and which one costs less and is faster/smoother?
Bionare vs maxwell? So you pick the only two recent architectures that could possibly make nvidia look decent and want me to take that comparison as a prediction on 20nm? Ridiculous!
It's pretty simple GK110 and hawaii are the respective high end 28nm parts, no?
How is it even close to relevant to compare bionare to nvidias current newest chip? GK110 = 560mm. Hawaii = 435 mm. GK110 5.2 GFLOPS SP Hawaii = 5.7 GFLOPS
AMD purposely made hawaii a compute card, look at the W9100.
If they want, they can grab the crown from nvidia any time they want.
More 4k benchmarks 290X vs 780ti with the fixed NV driver, the most recent WHQL release:
http://www.guru3d.com/articles_pages/asus_geforce_gtx_780_ti_matrix_review,16.html
and yet more:
http://www.guru3d.com/articles_pages/asus_geforce_gtx_780_ti_matrix_review,17.html
The 780ti has a commanding lead in all games tested @ 4k with the most recent WHQL. But the 290X is a better value since it is the cheaper card, and the mining inflation situation is starting to go away. And they're both fairly close in performance, and both very strong at 4k.
Hawaii is noticeably smaller, yet runs hotter and generally uses more power. Yet it's often able to keep up with GK110.
Hawaii is noticeably smaller, yet runs hotter and generally uses more power.
AMD is smaller but more DENSER, hence it runs hotter.
My prediction is that GM104 will be some 20-40% Faster than Hawaii XT, with 20nm AMD high end being some 20% faster than that.
Nvidia will then release full GM110 20nm at 0-10% faster than 20nm AMD high end and charge 50% more $.
AMD has always been denser.
Smaller and using more power is whats essential here; that's a double thermal jeopardy due to the way heat dissipation works. Temps are just consequence of this.
Sometimes 2d state on my 290 gets stuck at 3d clocks. Temps shoot at 90+C just by watching Youtube (proly stupid fan gets stuck at 20% too).
No other chip comes close to this.
Maybe your card is defective. ?
RMA it maybe..