You are making a basic mistake.The way some are normalizing Polaris 14nm area to 390 28nm area is not correct and leads to overoptimistic numbers. Not every unit of the GPU is going to scale linearly with process. In particular the physical I/O interfaces don't scale well and sometime they can actually scale negatively to compensate for different process characteristics (and inexperience..). At this point is more likely that Polaris matches a 390x in terms of transistor count, or perhaps it's s bit less, but it makes up for it with higher efficiency and better process.
Sent from my iPhone using Tapatalk
You are making a basic mistake.
390X = 512 bit and Polaris 10 = 256 bit. That is why you can effectively scale I/O using the same ratios. I/O interface is 1/2.
P10 should have the die size around the same as 7870 <- amazing chip besting maxwells gtx960 in recent titles in both performance and performance/watt, and it is even smaller than 960.
7870 is a 120W card. Around what we expect from P10, maybe even on the downside. Even when we assume P10 will take the same amount of juice as 7870, and we apply the 2.5x efficiency on top of it we arrive at:
True, but memory is by far the largest % of I/O area, so there is substantial scaling by going from 512 > 256 bit.And display?
Correct, but if several approaches indicate a similar end result, you can have greater confidence in the prediction.Scaling power with 2-2.5 announced efficency improvements to calculate P10 performance seem easier than going by the unknow area, 2-2.2 area scaling, and unknown perf/area improvements.
You are making a basic mistake.
390X = 512 bit and Polaris 10 = 256 bit. That is why you can effectively scale I/O using the same ratios. I/O interface is 1/2.
I don't put stock to any rumors/estimates since how wrong the predictions were for the entire year of 2015, but that is a really good point.Correct, but if several approaches indicate a similar end result, you can have greater confidence in the prediction.
its bigger than a 390x......almost 30mm2 bigger..
Yes, 384 bit on 7970 was bigger than 512 bit on 290X. That was a first design and no GDDR5X to help.FYI: the 512bit controller on hawai was done because it required less die space than a 256bit/384bit using higher speed memory.
You realize that to crush AMD and Polaris you must have stock in sufficient quantities to satisfy all demand? This is assuming that Polaris is non-competitive. Unless you have some insider knowledge and not sharing.
Do you really expect us to believe that Nvidia is rolling in stock ready to saturate the retailers? What a joke.
Unless you expect Nvidia to sell all of the 1070 cards they will ever sell and satisfy the market for this new card in June, a market will still exist. I wonder what they will produce after that miraculous feat? I guess that will release so many additional wafers for GP100 that we'll get big Pascal earlier.
Normally your posts are well reasoned, but the last couple days since the unveiling of 1080 & 1070 has seen some very, very strange posts by you. Very poorly thought out and reasoned positions. Almost as if a different person has emerged.
They aren't unreasonable at all.
I've never cheerleaded or loved AMD, so I hope you get that notion out of your head. Remember, I don't care. If AMD and Nvidia both die tomorrow, I won't lose a day of sleep over PC Gaming.
Nvidia will have preorders next gen for the Founders Edition mark my words. It's the intelligent thing to do. It lets people pledge their money early, which gamers LOVE to do, and means that those people will be more committed to buying an Nvidia product, and less likely to switch over or pay attention to an AMD launch.
This is Nvidia we're talking about, they make brilliant business moves regularly.
Just recently, rebranding the 680 into the 770, having the 780/780Ti performance sell horrendously inept cards like the 770 2GB.
Then, releasing the GTX 970 and selling it like hotcakes. Oh 3.5GB of VRAM? Oops. Slower than Hawaii? The architecture it was meant to blow away? Double Oops.
Releasing the 980 at $550, then releasing a 980Ti shortly after a $1000 Titan X for $650?
And now, Nvidia just getting warmed up. Founders edition is BRILLIANT. It gives Nvidia ANOTHER way to hyper segment people into charging them the most money possible.
So no, preorders is nowhere near an unintelligent thing for Nvidia to do. They WILL start doing it, mark my words.
I'm not even sure why you think a shortage would be a bad thing for the 1070 if Nvidia had preorders.
Have you seen Nvidia marketing?
"1070, so popular our waiting list of pledged people to purchase our card extends into July (Isn't that when Polaris is predicted to launch?)"
If you want to worry about the better product/deal, I think Polaris 10 is possible to be a good deal at 390x and under 250. Even at 250 it's a good deal and blows away the 1070 on price/performance.
Do you think the average gamer crowd is going to care though once they see the dominance of the 1070/1080? No, they'll wait for another Nvidia card to buy. People don't make logical purchasing decisions. They make emotional ones.
It's why Nvidia released the 1070 at the absurd price of $450, and a "lower" price of $380. Because people won't do the math to realize that a 1070 at even $380 is 1.9 times more expensive than a $200 Polaris 10, or 1.5 times more expensive than Polaris 10, while only delivering 25% more performance (390x is 80% performance of 980ti/1070). If AMD even wants a chance at selling Polaris 10, they need to drop the 390x performance comparisons. No gamer wants to hear that. They want to hear how it compares to a Titan X and sadly, AMD just doesn't sound great there.
Absolutely nothing in my post said you did. Lets be clear about that. On the other hand, I got a lot of laughs reading this, so thanks.They aren't unreasonable at all.
I've never cheerleaded or loved AMD, so I hope you get that notion out of your head. Remember, I don't care. If AMD and Nvidia both die tomorrow, I won't lose a day of sleep over PC Gaming.
Nvidia will have preorders next gen for the Founders Edition mark my words. It's the intelligent thing to do. It lets people pledge their money early, which gamers LOVE to do, and means that those people will be more committed to buying an Nvidia product, and less likely to switch over or pay attention to an AMD launch.
This is Nvidia we're talking about, they make brilliant business moves regularly.
Just recently, rebranding the 680 into the 770, having the 780/780Ti performance sell horrendously inept cards like the 770 2GB.
Then, releasing the GTX 970 and selling it like hotcakes. Oh 3.5GB of VRAM? Oops. Slower than Hawaii? The architecture it was meant to blow away? Double Oops.
Releasing the 980 at $550, then releasing a 980Ti shortly after a $1000 Titan X for $650?
And now, Nvidia just getting warmed up. Founders edition is BRILLIANT. It gives Nvidia ANOTHER way to hyper segment people into charging them the most money possible.
So no, preorders is nowhere near an unintelligent thing for Nvidia to do. They WILL start doing it, mark my words.
I'm not even sure why you think a shortage would be a bad thing for the 1070 if Nvidia had preorders.
Have you seen Nvidia marketing?
"1070, so popular our waiting list of pledged people to purchase our card extends into July (Isn't that when Polaris is predicted to launch?)"
If you want to worry about the better product/deal, I think Polaris 10 is possible to be a good deal at 390x and under 250. Even at 250 it's a good deal and blows away the 1070 on price/performance.
Do you think the average gamer crowd is going to care though once they see the dominance of the 1070/1080? No, they'll wait for another Nvidia card to buy. People don't make logical purchasing decisions. They make emotional ones.
It's why Nvidia released the 1070 at the absurd price of $450, and a "lower" price of $380. Because people won't do the math to realize that a 1070 at even $380 is 1.9 times more expensive than a $200 Polaris 10, or 1.5 times more expensive than Polaris 10, while only delivering 25% more performance (390x is 80% performance of 980ti/1070). If AMD even wants a chance at selling Polaris 10, they need to drop the 390x performance comparisons. No gamer wants to hear that. They want to hear how it compares to a Titan X and sadly, AMD just doesn't sound great there.
Absolutely nothing in my post said you did. Lets be clear about that. On the other hand, I got a lot of laughs reading this, so thanks.
It's pretty obvious that P10 has a better price/performance value than 1070.
That P10 CF will CRUSH the 1070SLI, 1080, and may even come close to 1080Sli given how horrendous SLI is (I don't even get it. Why do so many people use SLI? They do know it's bad right? Crossfiring 2 cheaper AMD cards is more effective).
3xP10 reigns supreme. I don't even think 3x1080 will beat out 3xP10.
Thing is, 3xP10 comes at the price of about 1 founders edition 1080.
Edit: Also, if Vega has UARCH improrvements, and isn't just HBM2, then I fail to see how Vega can be in the same stratosphere as Pascal.
Making claims about anything crushing anything else before we see a single benchmark (or official pricing information for P10) seems a bit premature
It is nowhere premature to say P10x2 will be cheaper than the 1080 founders edition, and CRUSH it in performance in games where CF scaling works. I think that's a given.
Just like how massively OCing a 980Ti won't make it faster than a 390x when CF scaling works....
CF Scaling is just good, and SLI scaling sucks. So P10s multi card configurations will be performance winners.
Unless you want to bet on Nvidia's 1070SLI/1080SLI being great performers at great prices....
Lol..
And do we know that SLI hasn't had a big overhaul? NVidia put a lot of emphasis on NVlink in the GP100 announcement, perhaps they used some of that tech to fix SLI.
And display?
And do we know that SLI hasn't had a big overhaul? NVidia put a lot of emphasis on NVlink in the GP100 announcement, perhaps they used some of that tech to fix SLI.
Jesus dude cheer up
Nvidia should have opened founders edition preorders on announcement. Polaris would have been crushed right away. I'm positive next gen thats how they one up what they did now
Or the fact Hawaii was designed with significant double precision performance for commercial compute tasks. Polaris is unlikely to have a lot of that functionality - you only need to look at Maxwell to see how much that helped with regards to gaming performance.
I am still hoping Polaris 10 can ~ Fury X for $299 but I'd rather stay cautious and assume it'll be ~ 390X for $249.
Or the fact Hawaii was designed with significant double precision performance for commercial compute tasks. Polaris is unlikely to have a lot of that functionality - you only need to look at Maxwell to see how much that helped with regards to gaming performance.