S/A: AMD kills Kaveri, Steamroller, Excavator

[pablo87]

Kabini 4C

http://wccftech.com/amd-richland-ap...-apu-radeon-hd-8000-series/#comment-716422150

These Kabini benchmarks are at Llano levels and basically close the file for big core because Trinity and any follow on are just not going to be competitive so AMD will be left with minute AM3+ and Server volumes to cover all that R&D and the GF WSA? Not going to happen.

Its also not difficult to imagine how much better they can make these with additional tweaking and resources, not to mention the # of permutations possible (id personally like to see a Kabini with an 8000 GPU and GDDR5).

Kind of explains why Lisa Su bought shares.

Rory Read just has to terminate the WSA with Global Foundries and give them some long term debt in lieu. (Rolls eyes).

PS - Llano benchmarks

http://www.notebookcheck.net/AMD-A-Series-A6-3420M-Notebook-Processor.61302.0.html

 

[mrmt]

Kind of explains why Lisa Su bought shares.http://www.notebookcheck.net/AMD-A-Series-A6-3420M-Notebook-Processor.61302.0.html

Don't be so quick. As an executive officer she has to own AMD shares in proportion to her salary.

Remember that she bought the stock when it was plunging, not when it was more or less stable.

 

[ShintaiDK]

http://www.menafn.com/menafn/74254d...o-Lay-off-Workers-Shanghai-RD-Center?src=main

First sign of who is getting the boot.

 

[Olikan]

http://www.menafn.com/menafn/74254d...o-Lay-off-Workers-Shanghai-RD-Center?src=main

First sign of who is getting the boot.

middle east north africa, financial network

dude, your deep web knowledge is insane

 

[nemesismk2]

AMD has issued a denial.

good

 

[pablo87]

Don't be so quick. As an executive officer she has to own AMD shares in proportion to her salary.

Remember that she bought the stock when it was plunging, not when it was more or less stable.

You are correct, I have to admit.

If these benchmarks are correct, the GPU benches are decent, and there are no major bugs, then this part potentially replaces both Brazos and Trinity by mid 2013.

Brazos is shipping 5MM per quarter, if Kabini raises those ASP's by $10, that's $50MM of additional profits. If it replaces 3MM of Trinity/LLano as well at wag $20 cost savings per part, that's another $60MM.

Not taking into account any marketshare gains from the part being so attractive to OEMs (common mb platform, integrated FCH, more competitive x86 performance), Kabini potentially represents additional profits of $440MM annually.

Not to mention future opportunities eg. higher performing GPU, micro-server, etc...So it is potentially a very positive development for AMD.

On the other hand, it assumes the BoD will make the right decision with respect to Global Foundries - if APUs at Gloflo were cancelled because of Kabini, then AMD's obligations under the WSA are most likely impossible to fulfill with the product that remains and the two companies with intertwined ownership and BoD would have to come to some sort of settlement. This is obviously completely outside of her control and probably RR's for that matter.

 

[mrmt]

Brazos is shipping 5MM per quarter, if Kabini raises those ASP's by $10, that's $50MM of additional profits. If it replaces 3MM of Trinity/LLano as well at wag $20 cost savings per part, that's another $60MM

I would not count on a 10 dollar increase on ASP more likely 5 and that for the next year only and I wouldn't bet on Kabini replacing Trinity until Q214 as AMD wouldn't launch Richland if they didn't expect the thing to last 12 months on the market.

You should also factor Silvermont/Airmont from Q413 pressing ASP from the bottom and Celeron/Pentium IVB/HSW from Q213 pressing from the top, I'd say that once those two factors are in place ASP will go down. All in all I expect very small profits from Kabini, ranging from a third to a half of what you expect, but is still a profit, which is better than anything is making in the last few years.

On the other hand, it assumes the BoD will make the right decision with respect to Global Foundries - if APUs at Gloflo were cancelled because of Kabini, then AMD's obligations under the WSA are most likely impossible to fulfill with the product that remains and the two companies with intertwined ownership and BoD would have to come to some sort of settlement. This is obviously completely outside of her control and probably RR's for that matter.

According to their annual report AMD must manufacture all its chips and APUs within GLF, with a few exceptions. Brazos is simply the most popular product from AMD line up, so I think they are already bending the limits of the WSA.

I'm sure GLF planned at least one operating fab of them entirely devoted to AMD and another running upgrades. So far GLF received more than 5 billion in cash and a huge part of it is dedicated to those AMD fabs.

Also AMD is GLF only significant customer. If they lose AMD they won't have anybody from the first tier chip designers with them. They would have to try to become a competitor to UMC and fight for the second and third tier designers. Not that this would be a bad thing, that's the right direction IMO, but one that they decided not to take. Keep in mind that the WSA is due 2024, so once you factor one billion in revenues per year for more 11 years you'll have a nice idea of how expensive GLF can sell the waiver. As they are really cash strapped, this decision is far beyond AMD BoD powers.

 

[Soleron]

Can we disregard Charlie already? Kaveri and Steamroller are very much alive according to the CES presentation. This is indefensible by him.

 

[mikk]

Charlie was wrong as usual.

 

[NTMBK]

Can we disregard Charlie already? Kaveri and Steamroller are very much alive according to the CES presentation. This is indefensible by him.

Sadly we can't- he was the first to break the news about Intel's on-package DRAM in their GT3e graphics. He does get genuine news stories quite often, it's just he's a bit, well, semiaccurate.

 

[Soleron]

Sadly we can't- he was the first to break the news about Intel's on-package DRAM in their GT3e graphics. He does get genuine news stories quite often, it's just he's a bit, well, semiaccurate.

OK can we just stop with the threads every time he predicts doom for a company? None of the doom has yet come to pass (on Nvidia, AMD, Intel graphics or Microsoft)

 

[Olikan]

Sadly we can't- he was the first to break the news about Intel's on-package DRAM in their GT3e graphics. He does get genuine news stories quite often, it's just he's a bit, well, semiaccurate.

charlie main problem is that, he don't just make a big news saying "i got it wrong"....hehe...he does that on the forums, but not to it's public...

in this case, charlie got wrong, it wasn't the core that was killed...but the chip...the octa-core steamroller based, that got replaced by trinity 2.0.. that's it

 

[mrmt]

OK can we just stop with the threads every time he predicts doom for a company? None of the doom has yet come to pass (on Nvidia, AMD, Intel graphics or Microsoft)

Maybe you should ask to AMD fired employees if they agree that none of the doom has yet come to pass.

 

[Centauri]

Because no company has ever fired anybody during a successful turnaround. Nope. Ever ever. Firings always mean bankruptcy.

 

[CHADBOGA]

charlie main problem is that, he don't just make a big news saying "i got it wrong"....hehe...he does that on the forums, but not to it's public...

in this case, charlie got wrong, it wasn't the core that was killed...but the chip...the octa-core steamroller based, that got replaced by trinity 2.0.. that's it

In the old days, one of Charlie's favourite tricks when he was exposed as being wrong was to say that he was right at the time, but something changed.

 

[MisterMac]

In the old days, one of Charlie's favourite tricks when he was exposed as being wrong was to say that he was right at the time, but something changed.

In fairness - could that not be what happened?

He got word of the Big Chip being killed - and assumed it was Big Core.

Or it was the Big Core entirely, but instead got mashed out to Big Chip lines.



People need to read SA like this forum - with salt and more logical deduction of all rumours currently floating.

So you like here - discuss all possible rumoured datapoints and make an "informed" "assessed geuss" - at how it will play out.

Charlie's rumours aren't untrue - they're just ALWAYS incomplete.
thats the problem - ours mind you, not his.

 

[AtenRa]

In fairness - could that not be what happened?

He got word of the Big Chip being killed - and assumed it was Big Core.

Or it was the Big Core entirely, but instead got mashed out to Big Chip lines.



People need to read SA like this forum - with salt and more logical deduction of all rumours currently floating.

So you like here - discuss all possible rumoured datapoints and make an "informed" "assessed geuss" - at how it will play out.

Charlie's rumours aren't untrue - they're just ALWAYS incomplete.
thats the problem - ours mind you, not his.

The news

Steamroller will not be released in 2013

The reporting,
AtenRa : AMD have said that SteamRoller will not be released in 2013 (2 hits)

Charlie : AMD have killed SteamRoller and all big cores and they are going bankrupt (1000000000 hits)

 

[MisterMac]

The news


The reporting,
AtenRa : AMD have said that SteamRoller will not be released in 2013 (2 hits)

Charlie : AMD have killed SteamRoller and all big cores and they are going bankrupt (1000000000 hits)

Wrong.
Mr. AMD Lover.

Spin more

The news:
FX big chip line is dead

Reporting:
Big Core line is completely dead - and only LP brazos\jaguar will be released.