HumanSmokeNice try with the trolling. The presentation and the slide are specifically Deep Learning orientated. Mixed precision - use of FP16 for object detection and recognition allied with a more comprehensive GPU to GPU/GPU to CPU interconnect. That is why the second half of the slide show the representation of four GPUs, because P100 - the focus of the presentation - has four NVLink interfaces.
I'm pretty sure only trolls and AMD fanboys try to make "Pascal 10X Maxwell" applicable across the board, rather than a the sole example it is intended for.

G33k2Fr34kEvery time I see a 10X or 100X from Nvidia, I know they're lying. Nvidia's PR is well known for shameless blatant lying. Only Nvidia fanboys come out to justify these ridiculous claims by Nvidia.

arbiterI think you confused Nvidia with AMD there.

G33k2Fr34kNope, I didn't. I remember back when Nvidia made the claim that their GTX280 was over 100X faster than Intel's best quad core CPU at the time running certain "CUDAH" tasks. Intel decided to verify these claims and the results of Intel's benchmarks were quite a bit different, to say the least. Intel found that their the GTX280 was was around 2.5 times faster than the Core i7 960 on average, and only in one scenario it was 14 times faster (probably a CUDAH app that uses old X87 instructions instead of SSE instructions).

G33k2Fr34kNope, I didn't. I remember back when Nvidia made the claim that their GTX280...

G33k2Fr34kNope, I didn't. I remember back when Nvidia made the claim that their GTX280 was over 100X faster than Intel's best quad core CPU at the time running certain "CUDAH" Intel decided to verify these claims and the results of Intel's benchmarks were quite a bit different, to say the least. Intel found that their the GTX280 was was around 2.5 times faster than the Core i7 960 on average, and only in one scenario it was 14 times faster (probably a CUDAH app that uses old X87 instructions instead of SSE instructions).

HumanSmokeAt least there WAS a GTX 280

medi01How much is that (in %)?

Ferrum MasterAh boney it is a AMD thread turned into intel vs nvidia war now? Calm down.

HumanSmokeafter all, at least two of the aforementioned are Disqus trollers of the highest order.

rruffI think you and many others are getting carried away again. The odds are very good that the full chip has 2048 shaders, and cut down versions will fill the space between this and Polaris 11. All indications are that this is a 480x so it will slot in well below the 1070, just as the 380x was well below the 970. So you'll need to wait until the fall before you see the cards that compete with the 1070 and 1080.

This is a smart move by AMD, because most likely their architecture will be inferior to Nvidia's but they will have several months where they have the best and newest midrange cards, since they'll be competing against Nvidia's Maxwell. How long has it been since that was the case?

FPS/$ will be at least as good as Nvidia's 1070. I predict Polaris 10 will match or slightly exceed 390/970 performance for $250 or less. It will be great if you are in that budget range.

MrGeniusWell there went 5 minutes of my life I'll never get back. Good thing I didn't waste the entire 7 minutes. Get to the point guy. Oh... you have no point. I see. Good bye! :rolleyes:

the54thvoidPut nem.. on my ignore list. At least he was easy to recognise from when he was plain old nem.

Still a class action Troll suit instigator.

dozenfuryOn these AMD-Nvidia debates that quickly descend into fanism, I'd suggest at least comparing apples to apples (MSRP to MSRP). Most people will wait for the MSRP models with better cooling, and AMD is sounding like they may not even have these Polaris cards in retail until August anyway. There's no requirement to pay a premium to be the absolute first to get the 1070/1080. And even on pure power tflops, Nvidia will always be ahead 1:1 on fps per tflop for most games due to better driver support.

Of course Polaris 10 won't compete with Pascal. It never was supposed to. It should compete with the R9 390 in a big way, and if AMD's new insulation technology along with other improvements allow for MUCH higher overclocks, then 32CUs might be all you need to reach near-980ti levels of performance. NEAR-980ti though. Let's not get ahead of ourselves.

ZoneDymoKids these days dont have any patients at all.....they need everything right now, right away.
The point was made at about 5:50, and its "just wait" before buying anything and he supported that argument with a long list of history of products released by every company and how waiting was often a better choice.
There ya go, now I do hope you learn some patients in the future.

Radeon lordPASCAL:
16nm maxwell with higher clock rates and
Some new Features like NVlink

MrGeniusWell as for patience I have more than I need(and then some). And I passed childhood well over 2.5 decades ago. His arguments were weak at best. Inaccurate and heavily biased for the most part. The Fury "overclocker's dream" comment seriously pissed me off, it's been OC to 1450/1000 with LN2(proving it truly is an overclocker's dream). The rest was just a bunch of "no duh" stupid and pointless BS for retards. So it most certainly didn't apply to me. That guy is a horrible "journalist"(term used in its broadest and most forgiving sense). He's a complete fucking moron IMO. I'll never make the mistake of watching another one of his vids again. That's all I meant by that.

Radeon lordPASCAL:
16nm maxwell with higher clock rates and
Some new Features like NVlink

nem..

...only trolls and AMD fanboys try to make "Pascal 10X Maxwell" applicable...

Bjorn_Of_IcelandYou mean a year later, they have now come up with something to compete against a 980ti?