Vega/Navi Rumors (Updated)

[crisium]

I do not believe that leak. That would be nice though.

 

[sm625]

It must be true, I'm going all in on AMD stock . If it is true AMD will have done an incredible job competing with both Intel and Nvidia with a fraction of either company's resources. Here's hoping.

It is generally not a good idea to go all in on any stock that just closed half the gap between it's recent highs and its 200 day moving average. Usually when that happens, it means the other half of that gap is going to close also, and you dont want to be around when it does.

 

[tential]

That's practically my life motto!

Ya pretty much how I was raised.
I'm not negative... I'd just rather be pleasantly surprised than letdown!

 

[Crumpet]

One Vega's specifications.

From a VEGA patch for linux

• case CHIP_VEGA10:
• adev->gfx.config.max_shader_engines = 4;
• adev->gfx.config.max_tile_pipes = 8; //??
• adev->gfx.config.max_cu_per_sh = 16;
• adev->gfx.config.max_sh_per_se = 1;
• adev->gfx.config.max_backends_per_se = 4;
• adev->gfx.config.max_texture_channel_caches = 16;
• adev->gfx.config.max_gprs = 256;
• adev->gfx.config.max_gs_threads = 32;
• adev->gfx.config.max_hw_contexts = 8;

https://lists.freedesktop.org/archives/amd-gfx/2017-March/006570.html

EDIT: FIJI Specs for Context:

• case CHIP_FIJI:
• adev->gfx.config.max_shader_engines = 4;
• adev->gfx.config.max_tile_pipes = 16;
• adev->gfx.config.max_cu_per_sh = 16;
• adev->gfx.config.max_sh_per_se = 1;
• adev->gfx.config.max_backends_per_se = 4;
• adev->gfx.config.max_texture_channel_caches = 816 (wurde von AMD später korrigiert);
• adev->gfx.config.max_gprs = 256;
• adev->gfx.config.max_gs_threads = 32;
• adev->gfx.config.max_hw_contexts = 8;

Found on reddit

 

[Bacon1]

One Vega's specifications.


Found on reddit

Yeah was posted last month: https://forum.beyond3d.com/threads/...d-vega-20-rumors-and-discussion.59649/page-64

 

[IEC]

That German says "was later corrected by AMD" re: the texture channel caches #, as it seems hilariously wrong. I believe it is supposed to be 16, not 816.

 

[tential]

AMD down 25% today... Why didn't I short at 14.... There was no way amd was going to sustain positive news. I should short now before Vega but no....

Sent from my Nexus 5X using Tapatalk

 

[tential]

AMD down 25% today... Why didn't I short at 14.... There was no way amd was going to sustain positive news. I should short now before Vega but no....

Sent from my Nexus 5X using Tapatalk

 

[Headfoot]

AMD down 25% today... Why didn't I short at 14.... There was no way amd was going to sustain positive news. I should short now before Vega but no....

Sent from my Nexus 5X using Tapatalk

leave the financial crap out of the thread. This is for Vega rumors.

 

[lobz]

@Markfw I'm not rebelling, just really curious, because you warn a lot of users here, but I don't understand, why is it OK to obviously provocate an entire community with passive aggressive trolling. I'm reffering to Mockingbird.

On topic: the leak with lower than 1070 performance while not fake, it's obviously not representative of the end product's performance, but sadly the other one seems to be an actual fake :| one can only hope it's not

 

[Malogeek]

but sadly the other one seems to be an actual fake

What makes it a fake compared to the much lower one?

 

[Markfw]

@Markfw I'm not rebelling, just really curious, because you warn a lot of users here, but I don't understand, why is it OK to obviously provocate an entire community with passive aggressive trolling. I'm reffering to Mockingbird.

On topic: the leak with lower than 1070 performance while not fake, it's obviously not representative of the end product's performance, but sadly the other one seems to be an actual fake :| one can only hope it's not

As a Mod post

Its not OK to troll. I may have to get more strict if this doesn't calm down.

 

[Crumpet]

Well now i'm even more confused.

Oh well, c'est la vie.

Back on topic, I agree the newer result is probably fake, but I also think that whilst it is at the far uppermost reaches of possibility, the result is plausible for a fully finalised Vega.

 

[lobz]

Well now i'm even more confused.

Oh well, c'est la vie.

I was wondering if I was the only one

 

[lobz]

What makes it a fake compared to the much lower one?

The missing link. Seems very improbable to me that AMD would only make sure that the good scores are deleted, and the hilariously bad ones can remain online..

 

[lobz]

As a Mod post

Its not OK to troll. I may have to get more strict if this doesn't calm down.

ummm... O.K.

 

[Markfw]

If anyone thinks a post deserves an infraction/warning, report it. If you don't like the moderation, start a thread in MD. Recently 2 posters were infracted, and one of them is getting a vacation., so there is moderation, and I am not letting people get away with much. If you desire, as I said, I can get more strict, and warn/infract for more minor violations.
Enough of this, or I will hand out infractions for calling out a mod.

 

[Malogeek]

The missing link. Seems very improbable to me that AMD would only make sure that the good scores are deleted, and the hilariously bad ones can remain online..

You're assuming they have control over that information? They can ask Futuremark but if they say no then they'd need to go through legal means of removing the scores. I doubt they'd care that much to remove any of the scores, if any of them are actually real.

 

[guskline]

I am looking forward to Big Vega to replace my 2 RX480s in CF with a single AMD card. In some benchmark results the twin 480s beat my GTX1080 but in most the GTX1080 still comes out on top AND it has tons of OCing headroom but did cost $699 new vs $480 total for both 480s.

Some of the speculation on overall performance gets a little crazy as the release dates nears. IF Big Vega is a bit faster than both 480s in CF I will be pleased.

I could get a GTX1080OC or even a GTX1080TI and best the 480s in CF but I would defeat the purpose of an "all AMD rig".

I patiently wait.

 

[Glo.]

https://fccid.io/MSIP-REM-ATI-102-D08601

 

[R0H1T]

So AMD needs a double pronged strategy with APUs for low end and full fledged high end notebook GPUs for the high end. For that to happen AMD have to be competitive with Nvidia in terms of perf/watt.

Don't think they need to, they simply need to be better or close to Intel (APU) + Nvidia (dGPU) to corner that part of that market, though the top end will be Nvidia for the foreseeable future because AMD don't have anything to replace that GTX 1070~1080 in a laptop, however the number of users who buy such a config is very limited.

 

[guskline]

https://fccid.io/MSIP-REM-ATI-102-D08601

Glo: If possible, could you explain this link more?

Is this the model number assigned to Vega?

 

[Crumpet]

Can anyone make more sense of this, I presume it means good things (and is hopefully not fake)

http://ranker.sisoftware.net/show_r...efdbe9dcebdaebcdbf82b294f194a999bfccf1c9&l=en

 

[Glo.]

Can anyone make more sense of this, I presume it means good things (and is hopefully not fake)

http://ranker.sisoftware.net/show_r...efdbe9dcebdaebcdbf82b294f194a999bfccf1c9&l=en

It means that the GPU was working on 1.15 GHz core clock, and SiSoft requires rewriting to utilize FP16 performance on Vega GPUs.

Glo: If possible, could you explain this link more?

Is this the model number assigned to Vega?

Well there is not much to think about a GPU from AMD, that has been registered for release.

If they were registered yesterday that means we are going to see the launch very soon.

 

[Crumpet]

It means that the GPU was working on 1.15 GHz core clock, and SiSoft requires rewriting to utilize FP16 performance on Vega GPUs.

Well there is not much to think about a GPU from AMD, that has been registered for release.

If they were registered yesterday that means we are going to see the launch very soon.

Ooh, and that was an MSI card by the looks of it, so that would mean that custom cooler brands will be available soon too.

GREAT (because my R9 290 is now flickering... D: )