AMD GDC2016 Thread

[el etro]

It is generally expected for the next gen flagship product to do better than the previous gen flagship. If the big Polaris somehow doesn't pull ahead of Fiji in nearly every categories, it would be a huge disappointment.


I really do not want to see another Bulldozer.

Real replacements to Tahiti, Hawaii and probably for Fiji arrive only next year with HBM2!

Polaris 11 is Cape Verde substitute, Polaris 10 is Pitcairn's.

 

[iiiankiii]

Real replacements to Tahiti, Hawaii and probably for Fiji arrive only next year with HBM2!

Polaris 11 is Cape Verde substitute, Polaris 10 is Pitcairn's.

Well, if that's the case, Nvidia better not release a 14nm card better than the GTX 980 TI during the same time frame as Polaris or AMD is, yet again, behind.

 

[Maverick177]

Real replacements to Tahiti, Hawaii and probably for Fiji arrive only next year with HBM2!

Polaris 11 is Cape Verde substitute, Polaris 10 is Pitcairn's.

GK104 at launch was much faster than the big Fermi though. Could you imagine if the 680 is slower than the 580?

Well, if that's the case, Nvidia better not release a 14nm card better than the GTX 980 TI during the same time frame as Polaris or AMD is, yet again, behind.

Exactly.

 

[Blitzvogel]

Yerli sounds drunk. Looks like he took Raja's introductory comments to heart

 

[Game_dev]

Well, if that's the case, Nvidia better not release a 14nm card better than the GTX 980 TI during the same time frame as Polaris or AMD is, yet again, behind.

That would be incredibly disappointing

 

[itsmydamnation]

Real replacements to Tahiti, Hawaii and probably for Fiji arrive only next year with HBM2!

Polaris 11 is Cape Verde substitute, Polaris 10 is Pitcairn's.

did they say 2017 in the stream? because the box is 2016........

That would be incredibly disappointing

depends on focus, if nv release DP compute heavy card first (they have contracts for this) then they are unlikely to have perf/watt crown (watch how magically perf watt doesn't matter anymore) and they wont have the laptop market. Unless you expect a full end to end release of 14nm GPU by end of Q3.........

 

[greatnoob]

So small Polaris' "mid-range" is 390 performance levels at a small power footprint or am I mistaken?

 

[PPB]

So small Polaris' "mid-range" is 390 performance levels at a small power footprint or am I mistaken?

I'm thinking about 110-130w for that Sku.

 

[S.H.O.D.A.N.]

That's what we're getting from what they've shown. Apply adjustments for PR speak as needed.

 

[Blitzvogel]

Damn AMD sucks at PR events.

Capsaicin? Where does AMD get the collective idea for shit like this? The Polaris name is ALREADY THERE. They need people who know how to speak, get the message across and get people excited. So painful to watch Lisa and Raja. And the lack of real Polaris news is jarring to the viewers who sat watching this crap for two hours.

Why can't they ever figure this out? Huddy is the only one that even feels like he has some presence. Hint: It's the accent.

 

[Elixer]

That live event was pretty bad.
No new info, except for the Radeon Pro Duo.
The rest? zzzzzzzzzzz

 

[Maverick177]

Yeah, AMD is just streets behind in terms of PR.

Also: https://twitter.com/PellyNV

 

[metalliax]

That live event was pretty bad.
No new info, except for the Radeon Pro Duo.
The rest? zzzzzzzzzzz

No new new info?

Did you miss the part where they put up an actual roadmap? Or demoed a Polaris GPU to the public? Or emphasized big gains to come in the future?

Sure the event wasn't amazing, however I wouldn't call it a bad production.

 

[PPB]

Yeah, AMD is just streets behind in terms of PR.

Also: https://twitter.com/PellyNV

Only shows every company has their own Roy Taylor.

 

[el etro]

did they say 2017 in the stream? because the box is 2016........

Re-checking it: Graph points 2017 begin for Vega.

Probably due to yields, they want to bake and sell the easier(and GDDR) chips first, when the process expertise increase in GF and HBM is ready they out the Vega series.

Making a 232mm^2 GPU match the best they have is already impressive and will do tons of goods to their mindshare. And i don't expect Nvidia being able to execute 16nm GPUs any better, unless they unleash the GP104 chip with GDDR instead of wait for HBM. Vega is just six months later than it...

 

[burninatortech4]

So 14nm is replacing mid range chips in 2016? True flagship replacement next year? Looks like my 390 is sticking around for a while.

 

[greatnoob]

I'm thinking about 110-130w for that Sku.

Seems like a pretty good upgrade for the 960/270 crowd, and after a year of 14nm production they might drop the 6-pin altogether! Transforming the loud, power-sipping reference 290 into a smaller and quieter 75w card sounds insane.

 

[Blitzvogel]

No new new info?

Did you miss the part where they put up an actual roadmap? Or demoed a Polaris GPU to the public? Or emphasized big gains to come in the future?

Sure the event wasn't amazing, however I wouldn't call it a bad production.

The production was bad. The amount of info given for the amount of yakking and filler was horrible. Engineers and team leads should not be marketing and giving presentations if it is not their forte. I could do a much better job getting the audience excited over this info, in a shorter amount of time. Speakers and presentations need an edge and an aura of presence. Steve Jobs had it, Jen-Hsun Huang certainly has more of it than anyone at AMD.

 

[sandorski]

Clearly they were just prepping for the really important releases of the vidcards a few months from now. Setting the stage for DX12 performance and the importance of VR.

 

[Blitzvogel]

Clearly they were just prepping for the really important releases of the vidcards a few months from now. Setting the stage for DX12 performance and the importance of VR.

30 Seconds of Polaris footage isn't much to go on when we don't know the specs of the GPU.

It was way too reminiscent of the 30 second debut of Llano, where they showed us some relatively uninteresting low-framerate footage of Aliens vs Predator. It was sort of interesting from a technical point of view, but it just didn't carry any sort of awe.

They should've made a montage with some good music, and both current and coming titles running on Polaris.

 

[Silverforce11]

This was really a VR showcase, all the partners and what's going on in the upcoming VR field. I'm not interested in VR, yet. But it was interesting to see it being applied for cinema and other fields besides gaming.

Polaris's main event will be for later prior to launch.

 

[iiiankiii]

No new new info?

Did you miss the part where they put up an actual roadmap? Or demoed a Polaris GPU to the public? Or emphasized big gains to come in the future?

Sure the event wasn't amazing, however I wouldn't call it a bad production.

I disagree. It was bad. It felt disjointed. These type of public PR events aren't about drowning the people with technical data. Focus on what makes the product amazing. Be more precise, to the point, and less technical. The technicality should be written on paper and handed out. I don't want to feel like I'm listening to a class lecture.

 

[Kenmitch]

I disagree. It was bad. It felt disjointed. These type of public PR events aren't about drowning the people with technical data. Focus on what makes the product amazing. Be more precise, to the point, and less technical. The technicality should be written on paper and handed out. I don't want to feel like I'm listening to a class lecture.

It wasn't a 1hr special on CBS. Wasn't it a room full of computer nerds/geeks?

 

[metalliax]

I disagree. It was bad. It felt disjointed. These type of public PR events aren't about drowning the people with technical data. Focus on what makes the product amazing. Be more precise, to the point, and less technical. The technicality should be written on paper and handed out. I don't want to feel like I'm listening to a class lecture.

You were not the audience for this. Gamers were not the audience. The audience is game developers @ GDC. AMD is simply trying to show the real developers what they've been doing over the past 6 months and what they plan to continue investing in.

 

[Bryf50]

I disagree. It was bad. It felt disjointed. These type of public PR events aren't about drowning the people with technical data. Focus on what makes the product amazing. Be more precise, to the point, and less technical. The technicality should be written on paper and handed out. I don't want to feel like I'm listening to a class lecture.

Game Developer Conference.