AMD CPU speculation... and expert conjecture

[jdwii]

@gunrave: amd eventually reduces cpu/apu prices which is why kaveri apus are as cheap as, if not cheaper and better than haswell/refresh core i3 cpus.

Thx for reply,
Yeah the 7850k is definitely Haswell refresh core i3 killer, but in the beginning the price are really not make sense, unless AMD know the their high end kaveri are outperform Intel counterpart and adjust the price to add the margin.

The last price-cut is, i think, the most attracting for kaveri to sale.

When the price get in to my country, i will buy it..haha

Right now the Fx 8320 is just 120$ as well

 

[szatkus]

Carrizo?
http://browser.primatelabs.com/geekbench3/compare/1279533?baseline=1279533

"Processor Codename Richland"

Seems legit.

Compare scores to other cpus though, as fearnot shows it nearly beats out kaveri at 1 ghz more, in fact compare with i5 2400s 2.5ghz and its close, real close.

What is turbo clock? L1i is too small, L2 too big.Nothing like Carrizo.

 

[juanrga]

Especially that 8350rocks said Carrizo is mobile only.

We will see Carrizo beyond mobile. In fact, AMD has already announced officially that Carrizo will be used in some "All-in-One systems".

"The 2015 desktop roadmap remains unpublished so far. AMD’s perception of a mobile-focused strategy would tend to suggest that the mobile comes first, with desktop following behind, although at this point it is unclear. A number of AMD’s marketing materials with this launch gave examples of the use of Carrizo and HSA for the prosumer, indicating that a desktop version should be announced in due course." - http://www.anandtech.com/show/8742/amd-announces-carrizo-and-carrizol-next-gen-apus-for-h1-2015

"Some rumors do point to the desktop variant being called Carrizo-D but AMD didn’t even mention about the desktop platform in the video but we already have some details as to what’s it all about." -http://wccftech.com/amd-officially-shows-carrizo-apu-sample-chip-carrizol-carrizo-apus-planned-1h-2015/

Also the fact that John Byrne discusses Carrizo as being a top-to-bottom product.

Carrizo is coming to mobile, desktop, and server...

 

[colinp]

76 frames per second. Wow, that's totally unplayable.

(A comment that makes no sense now that the troll has been banhammered)

 

[juanrga]

^^ It is not a question of playable/unplayable. The AMD card is providing 90--92% of the performance of the GTX 980. This small difference is statistically irrelevant, a simple drive update can close that.

 

[colinp]

Or a driver update from Nvidia could open up the gap further. Either way, it's playable.

 

[etayorius]

Especially that 8350rocks said Carrizo is mobile only.

We will see Carrizo beyond mobile. In fact, AMD has already announced officially that Carrizo will be used in some "All-in-One systems".

"The 2015 desktop roadmap remains unpublished so far. AMD’s perception of a mobile-focused strategy would tend to suggest that the mobile comes first, with desktop following behind, although at this point it is unclear. A number of AMD’s marketing materials with this launch gave examples of the use of Carrizo and HSA for the prosumer, indicating that a desktop version should be announced in due course." - http://www.anandtech.com/show/8742/amd-announces-carrizo-and-carrizol-next-gen-apus-for-h1-2015

"Some rumors do point to the desktop variant being called Carrizo-D but AMD didn’t even mention about the desktop platform in the video but we already have some details as to what’s it all about." -http://wccftech.com/amd-officially-shows-carrizo-apu-sample-chip-carrizol-carrizo-apus-planned-1h-2015/

Also the fact that John Byrne discusses Carrizo as being a top-to-bottom product.

Carrizo is coming to mobile, desktop, and server...

So it`s actually coming to Desktop? i thought it was not going to appear on Desktop.

 

[8350rocks]

Especially that 8350rocks said Carrizo is mobile only.

We will see Carrizo beyond mobile. In fact, AMD has already announced officially that Carrizo will be used in some "All-in-One systems".

"The 2015 desktop roadmap remains unpublished so far. AMD’s perception of a mobile-focused strategy would tend to suggest that the mobile comes first, with desktop following behind, although at this point it is unclear. A number of AMD’s marketing materials with this launch gave examples of the use of Carrizo and HSA for the prosumer, indicating that a desktop version should be announced in due course." - http://www.anandtech.com/show/8742/amd-announces-carrizo-and-carrizol-next-gen-apus-for-h1-2015

"Some rumors do point to the desktop variant being called Carrizo-D but AMD didn’t even mention about the desktop platform in the video but we already have some details as to what’s it all about." -http://wccftech.com/amd-officially-shows-carrizo-apu-sample-chip-carrizol-carrizo-apus-planned-1h-2015/

Also the fact that John Byrne discusses Carrizo as being a top-to-bottom product.

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

 

[de5_Roy]

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

looks like yet another failure from glofo. with kaveri, amd set an aggressive perf./power target that glofo seems to have failed to meet, which prompted the process switch and sr-B. with carrizo, amd was being more aggressive and ambitious. so glofo crapped out. that makes intel the only one, despite their struggles with 22nm, with a working shrunk node (i wonder if that came out right....), and finfet.. for what it's worth.

 

[gamerk316]

I'm going to continue to hold the belief, despite the short term cost savings, that spinning off GloFo was a massive strategic mistake for AMD.

 

[juanrga]

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

looks like yet another failure from glofo. with kaveri, amd set an aggressive perf./power target that glofo seems to have failed to meet, which prompted the process switch and sr-B. with carrizo, amd was being more aggressive and ambitious. so glofo crapped out. that makes intel the only one, despite their struggles with 22nm, with a working shrunk node (i wonder if that came out right....), and finfet.. for what it's worth.

No, it doesn't look like that. In the first place, both Carrizo and Carrizo Lite are built on 28nm, not 20nm. In the second place Glofo is not responsible because SOI didn't match the hype and the whole SOI industry collapsed (including IBM foundry business). Glofo migrating from 32nm SOI to bulk (28 and 20 nm) and next FinFET was the right move.

Third, what Intel calls 22nm is what rest of industry (TSMC, Glofo...) calls ~26nm. Intel fabs aren't so ahead as their marketing dep. pretends. Moreover, the 28SHP process used in Kaveri brings densities are not achievable by Intel 22FF process node.

Fourth, the 20nm node was never aimed to high-performance products, because the whole PC industry was declining (still is) and Glofo planned the 20nm node and successors for mobile and other products...

It is not random that AMD is introducing mobile products first, neither that Intel is doing the same with Broadwell Core M. The Broadwell-E chips for enthusiasts are delayed to 2016.

 

[juanrga]

Spoiler

Especially that 8350rocks said Carrizo is mobile only.

We will see Carrizo beyond mobile. In fact, AMD has already announced officially that Carrizo will be used in some "All-in-One systems".

"The 2015 desktop roadmap remains unpublished so far. AMD’s perception of a mobile-focused strategy would tend to suggest that the mobile comes first, with desktop following behind, although at this point it is unclear. A number of AMD’s marketing materials with this launch gave examples of the use of Carrizo and HSA for the prosumer, indicating that a desktop version should be announced in due course." - http://www.anandtech.com/show/8742/amd-announces-carrizo-and-carrizol-next-gen-apus-for-h1-2015

"Some rumors do point to the desktop variant being called Carrizo-D but AMD didn’t even mention about the desktop platform in the video but we already have some details as to what’s it all about." -http://wccftech.com/amd-officially-shows-carrizo-apu-sample-chip-carrizol-carrizo-apus-planned-1h-2015/

Also the fact that John Byrne discusses Carrizo as being a top-to-bottom product.

Carrizo is coming to mobile, desktop, and server...

So it`s actually coming to Desktop? i thought it was not going to appear on Desktop.

Yes. AMD has just announced officially Carrizo for All-In-Ones. There is another Carrizo desktop platform that is still unannounced. And then there are the rumors about a third line named Carrizo-D, although other sources say Carrizo-E.

 

[Cazalan]

I'm going to continue to hold the belief, despite the short term cost savings, that spinning off GloFo was a massive strategic mistake for AMD.

This wasn't done for short term cost savings. It was done to avoid bankruptcy. Even IBM can't make a fab profitable in 2014 so what chance did AMD have? GF backers have dumped roughly 20B into the fab to try and compete with TSMC. Money AMD had zero chance of securing.

AMDs biggest mistake has been breaking multiple contracts with GF that resulted in roughly 1.5B in penalties. Of which they still haven't recovered. Instead of 0.5B in debt they're over 2B in debt. Their interest payments on their debt are bleeding them dry and making many quarters that would have otherwise been profitable a loss.

 

[jdwii]

^^^ Look at global foundries lately the best thing Amd did is break contracts with them. Hope they change but right now they are a joke compared to Intel and TSMC.

 

[Cazalan]

^^^ Look at global foundries lately the best thing Amd did is break contracts with them. Hope they change but right now they are a joke compared to Intel and TSMC.

Actually they broke the contracts, paid the huge price, and now switched a lot of their production back to GF. The 28SHP is a more advanced process than TSMC 28nm. Even their new Tonga GPUs are made at GF now.

Granted the x86 exclusivity contract sucked but GF did take 3B (or was it 5B) of AMDs debt off their back. That's more than the entire profits from PS4/XBone combined over their lifetime and then some. From GF perspective they saved AMDs arse, and AMD bailed on them by giving more business to TSMC.

GF got smarter the second time around with IBM and practically told them to %&#. IBM ended up having to pay GF to take the fabs off their hands.

 

[szatkus]

^^^ Look at global foundries lately the best thing Amd did is break contracts with them. Hope they change but right now they are a joke compared to Intel and TSMC.

Actually they broke the contracts, paid the huge price, and now switched a lot of their production back to GF. The 28SHP is a more advanced process than TSMC 28nm. Even their new Tonga GPUs are made at GF now.

Granted the x86 exclusivity contract sucked but GF did take 3B (or was it 5B) of AMDs debt off their back. That's more than the entire profits from PS4/XBone combined over their lifetime and then some. From GF perspective they saved AMDs arse, and AMD bailed on them by giving more business to TSMC.

GF got smarter the second time around with IBM and practically told them to %&#. IBM ended up having to pay GF to take the fabs off their hands.

I'm pretty sure that Tonga is made at TSMC.

http://i.imgur.com/ABzFRth.jpg

 

[esrever]

AMD will probably shift production back to Glofo because its much cheaper for them, especially because they still have a set wafer agreement with glofo that they can't even make use of as of now. GPUs will likely stay at TMSC for the time being unless they can't get 16nm due to demand or something.

 

[anxiousinfusion]

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

Anything by the name Carrizo-D or Carrizo-E perchance?

 

[jdwii]

AMD will probably shift production back to Glofo because its much cheaper for them, especially because they still have a set wafer agreement with glofo that they can't even make use of as of now. GPUs will likely stay at TMSC for the time being unless they can't get 16nm due to demand or something.

Its probably cheaper since they're not always reliable tbh its a problem there all facing laltly

 

[jdwii]

AMD will probably shift production back to Glofo because its much cheaper for them, especially because they still have a set wafer agreement with glofo that they can't even make use of as of now. GPUs will likely stay at TMSC for the time being unless they can't get 16nm due to demand or something.

Its probably cheaper since they're not always reliable tbh its a problem there all facing laltly

 

[jdwii]

This one is for Juan happy seeing more arm products for servers
http://www.extremetech.com/computing/194701-qualcomm-will-enter-arm-server-market-with-major-partners-broad-solutions

 

[8350rocks]

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

looks like yet another failure from glofo. with kaveri, amd set an aggressive perf./power target that glofo seems to have failed to meet, which prompted the process switch and sr-B. with carrizo, amd was being more aggressive and ambitious. so glofo crapped out. that makes intel the only one, despite their struggles with 22nm, with a working shrunk node (i wonder if that came out right....), and finfet.. for what it's worth.

Hmm...last word I had, carrizo was ULP only. There will be something for DT to succeed kaveri. Essentially 20nm is not a good process for DT product, among other things...

Anything by the name Carrizo-D or Carrizo-E perchance?

Got clarification actually...they are going to do carrizo on 28nm across the board, so no CPU die shrink until new uarch. 20nm turned out so poor that they abandoned the idea of even using it for mobile at this point...so, not only will the DT parts be 28nm, but the ULP stuff will too now...

 

[juanrga]

This one is for Juan happy seeing more arm products for servers
http://www.extremetech.com/computing/194701-qualcomm-will-enter-arm-server-market-with-major-partners-broad-solutions

I already knew that notice, just as I also know that Cray has announced that will build supercomputers based in ARM SoCs made by another company. I know technical details and the performance of the 95W SoC...

 

[jed]

I'm going to continue to hold the belief, despite the short term cost savings, that spinning off GloFo was a massive strategic mistake for AMD.

+1 That's been said for years now,

 

[de5_Roy]

Qualcomm eyes server market
http://www.fudzilla.com/home/item/36377-qualcomm-eyes-server-market

AMD prolongs A-series software deal; price cuts still a work in progress
http://techreport.com/news/27400/amd-prolongs-a-series-software-deal-price-cuts-still-a-work-in-progress

semiaccurate's mantle test part 5
http://semiaccurate.com/2014/11/21/mantle-showdown-mantle-outperform-directx-11/

Got clarification actually...they are going to do carrizo on 28nm across the board, so no CPU die shrink until new uarch. 20nm turned out so poor that they abandoned the idea of even using it for mobile at this point...so, not only will the DT parts be 28nm, but the ULP stuff will too now...

if this is true, then glofo has screwed amd out of a potential step or three towards power efficiency advancement, die area, clockrate etc. i guess amd will launch zen and k12 on a new process and "hype" their "massive" power efficiency advantage over current apus and soc (seattle included). although it was already painfully obvious when amd had to put out a slide on "apu optimized process" b.s. lol