[TonyMacx86]Possible Polaris 12 GPU and Refreshed Polaris 10 spotted in macOS Sierra drivers.

[zinfamous]

If P11<P10, why whould P12>P10? makes no sense.

the numbering scheme for Polaris is designated in increasing order of the release schedule, as mentioned several times here and throughout the Polaris release since...when was that? June?

10 came first, then 11, then "12," maybe. It has nothing to do with die size.

supposedly, Vega 10 is small Vega and 11 is big Vega. So "a reverse," but not really because small Vega (10) is supposed to be released prior to big Vega (11).

 

[Phynaz]

GCN 1.2 Hawaii (R9 290X) was launched in October 2013,
Kaveri APU launched in January 2014, only two months after Hawaii.

Kaveri isn't GCN 1.2, it's 1.1 -
http://www.anandtech.com/show/7677/amd-kaveri-review-a8-7600-a10-7850k/4
https://en.wikipedia.org/wiki/List_...roprocessors#.22Kaveri.22_.282014.2C_28_nm.29

GCN 1.1 was released in March 2013. So it was 10 months between GPU and CPU.

 

[crisium]

Honestly, it's hard to make heads or tails of this. It's probably not a refresh of P11 because they also confirm a refresh of P10 that's named differently

This is why P12 is likely a new chip. If their revision to P10 is called P10 XT2, then why would a revision to P11 be called P12?

I don't think it's something smaller than P11. Is it worth R&D and manufacture of something smaller than the cutdown P11? Who on earth is the customer?

In-between P10-P11 is possible, but I'm not sure what the point is. A fullsize, smaller chip (say if P12 is 1536SPs) would compete better with GP107 (larger P10 costs a lot more, slower P11 is just that - slower), but is it really worth it for the $120-$160 segment? Really?

The linked in leak confirmed (using that word lightly) 4096SP Vega chip. We still do not know if that's big or small Vega. If it is small, then there is plenty of room in between P10 and Small Vega for Polaris 12.

2304SP 32 ROP 256-bit. 4096SP 64 ROP (presumably) HBM2, and a 3072SP 64 ROP 384-bit Polaris 12?

That is simply my guess. 2816SP Hawaii and 2048SP Tahiti/Tonga coexisted. 3072/2304 is a very similar ratio.

Scorpio GPU could be a sign of that, but not necessarily. IIRC, PS4 uses a 1280SP (cut down to 1152) custom chip that has 8 ACE's, essentially making it a GCN 2 Pitcairn. But AMD instead milked the immortal Pitcairn and never released something like this.

But... they may be more likely to create a similar dGPU this time around because there's nothing old to milk if there's a big gap between P10 and small Vega since this is a new node. Not gonna keep making Fiji, obviously, and that's the only AMD chip that will fit between P10 and small Vega performance (if small Vega is 4096SPs).

 

[happy medium]

If I was AMD, I would have a second cut down Vega 10.
Vega 10/Fury X smack in between the 1080 and Titan.
Vega 11/cut fury in between the 1080 and 1070
Vega 12/ futher cut fury in between the 480 and and Vega 11

That's the traditional way they do it.

 

[Glo.]

The linked in leak confirmed (using that word lightly) 4096SP Vega chip. We still do not know if that's big or small Vega. If it is small, then there is plenty of room in between P10 and Small Vega for Polaris 12.

2304SP 32 ROP 256-bit. 4096SP 64 ROP (presumably) HBM2, and a 3072SP 64 ROP 384-bit Polaris 12?

GDDR5X or HBM2 for this.

 

[Gigamaxx]

So far, I've only seen DDR3, DDR4, GDDR5, HBM, PCI in the memory section of the AMD4100kext. This is the kext dedicated to Fiji, Tonga, Polaris, and now Vega chipsets. No GDDR5, no HBM2 listed (visibly) as of yet.

Maybe we get a Christmas card launch with HBM memory on a Polaris card? Dec. 13th will be interesting.

 

[AtenRa]

Kaveri isn't GCN 1.2, it's 1.1 -
http://www.anandtech.com/show/7677/amd-kaveri-review-a8-7600-a10-7850k/4
https://en.wikipedia.org/wiki/List_...roprocessors#.22Kaveri.22_.282014.2C_28_nm.29

GCN 1.1 was released in March 2013. So it was 10 months between GPU and CPU.

Yea my bad, Hawaii and Kaveri are GCN 1.1 not 1.2

Hawaii (GCN 1.1) launched October 2013

Kaveri (GCN 1.1) launched January 2014

Only two months apart.

 

[Phynaz]

Yea my bad, Hawaii and Kaveri are GCN 1.1 not 1.2

Hawaii (GCN 1.1) launched October 2013

Kaveri (GCN 1.1) launched January 2014

Only two months apart.

Hawaii wasn't a new architecture, it was a beefed up Tahiti. We're back at 10 months

 

[AtenRa]

Hawaii wasn't a new architecture, it was a beefed up Tahiti. We're back at 10 months

by the same logic Polaris is not a new architecture, its a beefed up Tahiti.

 

[Phynaz]

by the same logic Polaris is not a new architecture, its a beefed up Tahiti.

Polaris is GCN 4, a new architecture. See the difference?

Look, I get you are holding out hope and you're trying like heck to bend history to fit your world view. Instead of trying to (and losing) to win an argument based upon semantics, just make your next post link to any kind of authoritative source that Zen APUs contain Vega GPUs. Just one. How about leaked slide deck? A quote from AMD? Any evidence at all?

 

[AtenRa]

Polaris is GCN 4, a new architecture. See the difference?

What is the difference that make GCN 1.4 Polaris a new architecture and GCN 1.1 Hawaii a beefed up GCN 1.0 Tahiti ???

Look, I get you are holding out hope and you're trying like heck to bend history to fit your world view. Instead of trying to (and losing) to win an argument based upon semantics, just make your next post link to any kind of authoritative source that Zen APUs contain Vega GPUs. Just one. How about leaked slide deck? A quote from AMD? Any evidence at all?

I havent mentioned ZEN APUs in our conversation, i was only pointing out that AMD have in the past used an APU with an IGPU tech only 2 months after a dGPU was launched.

 

[ElFenix]

Polaris is GCN 4, a new architecture. See the difference?

Look, I get you are holding out hope and you're trying like heck to bend history to fit your world view. Instead of trying to (and losing) to win an argument based upon semantics, just make your next post link to any kind of authoritative source that Zen APUs contain Vega GPUs. Just one. How about leaked slide deck? A quote from AMD? Any evidence at all?

who gives a #$%&?

 

[Glo.]

Polaris is GCN 4, a new architecture. See the difference?

Look, I get you are holding out hope and you're trying like heck to bend history to fit your world view. Instead of trying to (and losing) to win an argument based upon semantics, just make your next post link to any kind of authoritative source that Zen APUs contain Vega GPUs. Just one. How about leaked slide deck? A quote from AMD? Any evidence at all?

Now, go back to the link that I have brought to you for the same question, and look at what is common denominator.

http://videocardz.com/62250/amd-vega10-and-vega11-gpus-spotted-in-opencl-driver

SI: TAHITI
CI / GFX7: MILOS, KRYPTOS, HAWAII, NEVIS, PENNAR, BONAIRE, Kabini
VI / GFX8: ICELAND, TONGA, CARRIZO, BERMUDA, racerx, FIJI
GFX81: AMUR, STONEY, ELLESMERE, DERECHO
GFX9: GREENLAND, RAVEN1X, VEGA10, VEGA11

I will put to you this way. There are two possibilities, that there is wishful thinking. AtenRa's for Zen APUs to use Vega GPUs. Or your's wishful thinking for Zen APUs to not use Vega GPUs.

If Zen APUs will use HBM in any way, shape or form, they have to use Vega architecture.

 

[ElFenix]

nm

 

[Cookie Monster]

Wouldn't be too bad to have a beefed up Polaris with 48CUs to compete against the GTX1070 while challenging the GTX1080 with a dual P10 solution while VEGA10 is released.

 

[Glo.]

https://semiaccurate.com/forums/showthread.php?p=252525#post252525

Charlie many months ago have said that there are 3 GPUs from Polaris lineup. Now things are getting interesting.

What if Polaris was supposed to compete with Pascal lineup, and Vega to compete with Volta/GP100 architecture derived for consumer?

 

[Gigamaxx]

Falcon Ridge is previous name for Thunderbolt 2 controller, from Intel.

Raven thunderbolt 3?

 

[Headfoot]

If there is a third polaris die, I doubt it would be bigger. More likely its the 64-bit ultra low end die, since the super-low end has not been updated by either camp for a very long time.

 

[Glo.]

Raven thunderbolt 3?

Raven Ridge is the codename for AMD Zen APUs.

 

[AtenRa]

If there is a third polaris die, I doubt it would be bigger. More likely its the 64-bit ultra low end die, since the super-low end has not been updated by either camp for a very long time.

With APUs having faster iGPUs, I dont see why they should release anything smaller than P11.

 

[Headfoot]

With APUs having faster iGPUs, I dont see why they should release anything smaller than P11.

I cant imagine they wont offer something here still. Sure iGPUs fill this niche, but I doubt they'll totally abandon it especially considering they would want to sell this sort of chip for Intel CPU based OEM builds.

 

[AtenRa]

I cant imagine they wont offer something here still. Sure iGPUs fill this niche, but I doubt they'll totally abandon it especially considering they would want to sell this sort of chip for Intel CPU based OEM builds.

Well they already have faster Cards than Intel iGPU at sub $100 (R4/R5 etc), why should they even care about this segment anymore and spend time and resources for it when they still sell those 28nm small dies. Dont forget that P11 is already small at 123mm2.