Polaris 10 Delayed? (Rumour)

[Sweepr]

Polaris 10 just got pushed back to October.

Link 1

What I can tell you, and this is only part of the reason, is that they need the time to get their GPU up to speed. It's failing validation above about 850 MHz.

Link 2

I've provided plenty of references for my credibility. I stand by AMD releasing Polaris in October.

Link 3

He's not wrong, the fabrication process just needs a little maturing. Polaris 10 is supposed to run at close to double the speed that it currently is. This combined with another reason equals a delay.

As for taking a ban if I'm wrong, so far I've been correct on GTX Titan performance, GTX Titan pricing, Fury X frequencies on stock voltage, lack of Fury X scaling with voltage, BIOS updates to block BCLK overclocking on non-K series processors, Pascal TDP, GTX 1080 memory frequency, GTX 1080 GPU frequency (OK I have been vague on that but I did say you'd see CPU-like frequencies on LN2), GTX 1080 2-way SLI performance vs GTX 980 TI 4-way SLI performance (watch HWBOT the day the embargo lifts), as well as many other things which were on my now defunct website such as Bulldozer performance not being the INTEL destroyer the world believed it would be, etc. You tell me.

Link 4

And then we have this piece from Nordic Hardware suggesting there will be no cards at Computex. Meanwhile the Macau event could be under embargo.

Even more frustrating is said to be the fact that AMD does not seem ready to show off some new graphics card public during Computex in early June. While the company is a closed press event in Macau on 26th May, the information conveyed there to be under embargo and not get published until later in June. AMD's partners hoped to showcase new graphics card models at the Computex trade show stands so far with the gag.

http://www.nordichardware.se/nyhete...osakerhet-hos-amd-partners-infor-polaris.html

I've reached out to AMD for clarification and will edit when I hear back, but for now - rumor has it, as well as backed up by a rather large AMD partner, that there will be no Polaris-based video cards from AMD partners at Computex. I also thought I'd talk with NVIDIA, who said that there will be a crap load of custom GTX 1080s at Computex, which is exciting - especially with things kicking off two weeks from now.

www.tweaktown.com/news/52056/rumor-amd-partners-showing-polaris-cards-computex/index.html

Polaris developer diary ② - May 4, 2016:

- Ellesmere XT still belongs to A0 chip testing phase

- we expect the third week of may to provide A1 official version of the chip for testing

- RTG has now issued a PCB design reference suggests, the various AIB has entered card pre-production stage, relatively speaking, Baffin PCB's proposal is very short and simple, 4-layer PCB + single fan on can, but also shorter than the Nano

- Ellesmere of PCB recommendations are slightly longer than the Nano, but as long as the 6-layer PCB + single fan

http://tieba.baidu.com/p/4523923961?qq-pf-to=pcqq.group

As of a week ago, Ellesmere XT was still A0, while A1 was expected in the third week of May. Do they need another stepping to get the clockspeeds they're planning?

If Polaris 10 is delayed (remmember official word is still mid-2016!), would this affect your purchase decisions?

 

[tential]

Is this a global foundries release? Because if so I don't expect anything of this chip. Gf has never delivered.

 

[ShintaiDK]

Is this a global foundries release? Because if so I don't expect anything of this chip. Gf has never delivered.

It should be produced at Glofo.

It seems 14LPP is plagued by parametric yield issues if true. Unless its directly a design issue.

And that somewhat aligns up with the leaks we have seen clocked around 800Mhz. No Polaris at Computex is another.

 

[Magee_MC]

Link 1



Link 2



Link 3



Link 4

And then we have this piece from Nordic Hardware suggesting there will be no cards at Computex. Meanwhile the Macau event could be under embargo.



http://www.nordichardware.se/nyhete...osakerhet-hos-amd-partners-infor-polaris.html



www.tweaktown.com/news/52056/rumor-amd-partners-showing-polaris-cards-computex/index.html



http://tieba.baidu.com/p/4523923961?qq-pf-to=pcqq.group

As of a week ago, Ellesmere XT was still A0, while A1 was expected in the third week of May. Do they need another stepping to get the clockspeeds they're planning?

If Polaris 10 is delayed (remmember official word is still mid-2016!), would this affect your purchase decisions?

Well it certainly makes me feel better that I pulled the trigger on a new $330 390X the other day to replace my wonky 290.

 

[Mopetar]

If that's even half true AMD may as well just find a hole to die in as that would be devistating. Seems to fly in the face of the Vega in October rumor and everything else AMD has said up to this point.

 

[tential]

It should be produced at Glofo.

It seems 14LPP is plagued by parametric yield issues if true.

All I know is gloflo never delivers so expecting Polaris on time or to even hold up to expectations is a mystery.
Polaris is a great chip in theory and I like the theory of it.

The reality is gloflo sucks at execution. Amd sucks at execution. Combine the two and while I want Polaris 10, I don't expect it's launch to go without a hiccup. Amd sucks at releasing products whole Nvidia launches are so buttery smooth I'd eat off their launch floors.

 

[Rvenger]

Jeez, just release them now at the slower speed and then release a Ghz edition in October. Get something mainstream out the door at least.

 

[DDH]

The anecdotal evidence in this thread is overwhelming. Combined with the anonymous second hand information from an anonymous source deep in the janitorial department at AMD, Im hopping on the AMD is dead bandwagon before it gets too crowded

 

[flopper]

an attempt to make amd comment

 

[monstercameron]

I wonder if they are just seeding false information to keep buyers on their toes?

 

[tential]

The anecdotal evidence in this thread is overwhelming. Combined with the anonymous second hand information from an anonymous source deep in the janitorial department at AMD, Im hopping on the AMD is dead bandwagon before it gets too crowded

If Polaris launch is bungled that's 3 launches in a row amd has messed up. I know the amd is dead bandwagon is lame, but I mean, it's a massive uphill battle if amd screws this launch up. I've already seen amd faithful leave the amd camp after Fiji I think even more will leave after Polaris if that's a failed launch.

 

[amenx]

Jeez, just release them now at the slower speed and then release a Ghz edition in October. Get something mainstream out the door at least.

If performance is far below expectations it could seriously damage their brand. And if they want to sell low performing cards that were meant to be a higher tier up, it could hit their margins pretty bad.

 

[bystander36]

Jeez, just release them now at the slower speed and then release a Ghz edition in October. Get something mainstream out the door at least.

No. If it's running at half the expected clocks, its performance will not be good enough to sell.

 

[Bacon1]

Didn't your thread about this get closed earlier today? You sure love the doom and gloom on Polaris.

 

[cytg111]

We all knew/know the odds with glofo, hardly a surprise if true. Saddening (all tech screwups are) but not surprising.
Still only conjecture though.

 

[Azix]

combined with the rumor vega was moved to october... makes no sense. After months of having the chip ready and demonstrated..

 

[Flapdrol1337]

They were having some sort of press event the 26th right?

 

[swilli89]

combined with the rumor vega was moved to october... makes no sense. After months of having the chip ready and demonstrated..

Yes.. they had working Polaris 11 in December of last year showing good performance. With all of the testing and validation controls they must have undergone already for PS4 Neo, Nintendo NX, Apple macbook design wins, I'd be utterly stunned if they don't have silicon ready to ship right now.

Well... we will know one way or another sooner or later. The plot thickens..

 

[linkgoron]

You've already had one thread locked... I'm not saying the story is false, but I agree with this post, threads based on forum posts should probably wait a day or two, to see if things are actually true.

Using Nordic and TweakTown looks like you have two sources, but Tweaktown is obviously just reporting from the Nordic (as they clearly state), and unless I'm misunderstanding something, the Baidu link is also a forum post.

So you have one poster with multiple posts, another forum poster somewhere else, and Nordic.

Anyway, I think that we should just have one big thread about Polaris, we're getting too many threads for every half-rumor, lots of flaming and trolling, with some posters clearly enjoying themselves.

 

[Flapdrol1337]

Yes.. they had working Polaris 11 in December of last year showing good performance.

They showed it vs a gtx 950. But yeah, seems weird they wouldn't have something ready to go by now.

 

[Sweepr]

You've already had one thread locked...

Different rumour now. 'Vega in October' is also based on forum posts, but for some reason there's no rumour in the title.

 

[linkgoron]

Different rumour now. 'Vega in October' is also based on forum posts, but for some reason there's no rumour in the title.

Don't you agree with your own post that people shouldn't open threads based on forum posts, and that such threads should be closed?

Also, the vega in October thread has "videorcardz" in the title, it's pretty clear that it's a rumor.

 

[Rvenger]

Reading the "Source" of this information, I don't believe this rumor. OCN member with 16 rep. Eh...

 

[Bryf50]

Polaris 10 will be clocked at 2.5ghz will be 20% faster than a GTX 1080 and will be on store shelves this Friday. You heard it here first, please alert the journalists.

 

[monstercameron]

Polaris 10 will be clocked at 2.5ghz will be 20% faster than a GTX 1080 and will be on store shelves Friday. You heard it here first, please alert the journalists.

source? :sneaky:

P.S. I got the joke...