AMD Pulls Radeon "Vega" Launch to October

[btarunr]

In the wake of NVIDIA's GeForce GTX 1080 and GTX 1070 graphics cards, which if live up to their launch marketing, could render AMD's high-end lineup woefully outperformed, AMD reportedly decided to pull the launch of its next big silicon, Vega10, from its scheduled early-2017 launch, to October 2016. Vega10 is a successor to "Grenada," and will be built on the 5th generation Graphics CoreNext architecture (codenamed "Vega").

Vega10 will be a multi-chip module, and feature HBM2 memory. The 14 nm architecture will feature higher performance/Watt than even the upcoming "Polaris" architecture. "Vega10" isn't a successor to "Fiji," though. That honor is reserved for "Vega11." It is speculated that Vega10 will feature 4096 stream processors, and will power graphics cards that compete with the GTX 1080 and GTX 1070. Vega11, on the other hand, is expected to feature 6144 stream processors, and could take on the bigger GP100-based SKUs. Both Vega10 and Vega11 will feature 4096-bit HBM2 memory interfaces, but could differ in standard memory sizes (think 8 GB vs. 16 GB).



View at TechPowerUp Main Site

 

[Eroticus]

BF1 Edition is coming =].

 

[RCoon]

Guess that means NVidia won't respond with the 1080ti until after October. Gotta wait even longer

 

[Nokiron]

What's with the GP102 in the source? 4500 shaders? That's more than the GP100.

Same with Vega 11, should that not be 4096 shaders and Vega 10 with 3584?

There is a lot of weird information here.

 

[LightningJR]

huh. doesn't sound good to me.

 

[MxPhenom 216]

Guess that means NVidia won't respond with the 1080ti until after October. Gotta wait even longer

I didnt expect them to release the 1080Ti till like 6 months after the 1080 release. Why expect it to be sooner than that?

 

[the54thvoid]

So if HBM2 isn't a supply problem, why not release the full chip in October? For all those that bemoan Nvidia for holding out high end products, this (if true) is exactly what AMD are doing. FTR, that's fine but a little less hypocrisy all around would be nice to see in this thread's upcoming posts. See, AMD will only release a cut down version because it dare not play it's full hand before GP100. To do that would be bad judgement, so instead they release a competitor to the 1080 (according to OP) and wait to see what Nvidia bring.
It's all cat and mouse so it'd be really mature if people recognised that is what business is all about. Our impatient needs dont really matter to AMD or Nvidia. If true, this is a good play by AMD, to release a crippled version of Vega.

Then a gain, could be FUD or stealth PR.

 

[RejZoR]

I was in dilemma of selling GTX 980 for one of these, but from the looks of it, I'll be using it for quite a lot more time, potentially even skipping GTX 1000 and R9 400 series.

 

[Ubersonic]

If Nvidia's new cards have caused AMD to accelerate the launch of Vega so much does this mean they think Polaris will be outmatched?

 

[the54thvoid]

If Nvidia's new cards have caused AMD to accelerate the launch of Vega so much does this mean they think Polaris will be outmatched?

Almost every source thus far (apart from 'fans') has suggested that Polaris will not better 980ti. Given 1080 is suggested to beat 980ti, then yes, they expect to be outmatched.

 

[Eroticus]

If Nvidia's new cards have caused AMD to accelerate the launch of Vega so much does this mean they think Polaris will be outmatched?

I think AMD already said Polaris is low-mid range cards.

 

[The Quim Reaper]

If Nvidia's new cards have caused AMD to accelerate the launch of Vega so much does this mean they think Polaris will be outmatched?

Polaris cards are nothing more than 390X performance brought down to $300/£200 price levels.

Vega is AMD's answer to the 1070 & 1080

 

[RCoon]

Why expect it to be sooner than that?

Cause I NEED IT NOW

 

[NC37]

Well that answers what I assumed about Polaris. It's just midrange. There is no competitor for what nVidia is bringing to the table. Which gives nVidia quite a big lead seeing it is hitting on the back to school times.

 

[john_]

Almost every source thus far (apart from 'fans') has suggested that Polaris will not better 980ti. Given 1080 is suggested to beat 980ti, then yes, they expect to be outmatched.

The rumors say that Polaris is a 230mm chip. So it can't fight 1070/1080 that are about 340(?)mm chips. And if 1070 is almost as fast as 980Ti, at least in resolutions where that 256bit data bus is not a problem, then Polaris can't match that performance of 1070, therefor the performance of a 980Ti.

AMD probably didn't expected Nvidia to offer performance. They where probably expecting a $400-$450 1070 that will offer performance closer to GTX 980 than 980Ti, added with better efficiency and better DX12 support. That was my guess, by I don't get paid thousand of dollars to predict how my competitor will move. The stuff that should predict that, failed. So now they are in a hurry to save that 20-24% of their market share.

 

[Enterprise24]

Can Vega OC compete with GP100 @ 2.1-2.2Ghz.
I don't like Fury Series because very low OC potential. Hope that AMD will fix this.

 

[EdInk]

AMD probably didn't expected Nvidia to offer performance. They where probably expecting a $400-$450 1070 that will offer performance closer to GTX 980 than 980Ti, added with better efficiency and better DX12 support.

I think they weren't expecting the price points Nvidia stated..as they already mentioned polaris will be a mainstream card priced so could potentially be around the $250 - $350 mark (470,480) and then they got shot in the face by 1070 prices at $380 giving Titan X performance.

Now it's time for damage control with all this reactive actions

 

[bug]

Oh, a rushed product. That's going to end well...

 

[medi01]

Almost every source thus far (apart from 'fans') has suggested that Polaris will not better 980ti. Given 1080 is suggested to beat 980ti, then yes, they expect to be outmatched.

Indeed. Although, nobody has expected them to beat 1080 anyway.
What I still wonder is, whether they can take on 1070 with higher clocks (that quantum break demo was 800Mhz)
Oh well, we'll see it soon.

 

[jigar2speed]

Need more information on multiple chip module. Does the OS treat 2 modules as 1 GPU chip ?

 

[btarunr]

Need more information on multiple chip module. Does the OS treat 2 modules as 1 GPU chip ?

Memory on the package, just like "Fiji" and GP100. That's what makes this MCM.

Any two chips sharing a package constitutes an MCM.

 

[Xajel]

Need more information on multiple chip module. Does the OS treat 2 modules as 1 GPU chip ?

Man it's like Fury, GPU + HBM memory dies are in the same packaging...

http://cdn.wccftech.com/wp-content/uploads/2015/06/AMD-Fiji-GPU-High-Resolution-Shot.jpg

Only Vega should come with HBM 2 rather than HBM 1 on current Fury cards... so it will allow 8GB vs 4GB limit on HBM... that's why Fury cards only have 4GB of video memory ( except the Duo Pro which has 2x 4GB configuration duo to two GPU's there )

 

[buggalugs]

AMD probably didn't expected Nvidia to offer performance. They where probably expecting a $400-$450 1070 that will offer performance closer to GTX 980 than 980Ti, added with better efficiency and better DX12 support. That was my guess, by I don't get paid thousand of dollars to predict how my competitor will move. The stuff that should predict that, failed. So now they are in a hurry to save that 20-24% of their market share.

The reference 1070 is $449 and with limited availability and no competition for a while, expect to pay over $500 , probably around $549 for a 1070 at least until October. Same with the 1080, the reference card is $699, but will be more like $749 or more for months.

I really dont think AMD have been caught out, Everyone in the industry has expected big performance gains with 16nm and 14nm. Die shrinks usually at least double performance from previous gen. Both companies have been slow getting to 16/14 nm and AMD are dealing with a new fab, and new technology like HBM/2. Actually the 1080/1070 is slower than I expected, in reality they are only slightly faster than last gen's highend like 980Ti and TitanX.

Both companies would be planning for a refresh too, You can be sure the 1080 and 1070 isnt a maxed out 16nm GPU, they always leave a little in the tank for Ti versions, then a refresh or 2 to milk the market and get their moneys worth.

 

[TheGuruStud]

In another couple years we can bring back, "It's over 9,000!".

 

[Relayer]

I would guess that it's a lack of HBM2 that's delaying the launch from now. Maybe HBM2 is yielding better than was predicted? Or they are now willing to launch with very limited supply?

IF Vega can smash GP104 and they can get the jump on big Pascal they can likely charge a premium to mitigate the supply issue.

If that's the case and they are going to charge a premium, then they better make sure it's a premium product.