[TechEye] GlobalFoundries to buy IBM Semi

[NostaSeronx]

Referring to Google isn't a source...

I'm referring to articles that google detects. I'm not using google as a source but the articles found by google.

There is a huge difference in how fast AMD was going before 2011 and how slow AMD is going after 2011. None of the issue were GlobalFoundries fault the sole blame of delays are AMD.

To dive into the difference:
28nm SHP was going to be used for four GPUs: Cape Verdi, Bali, Tiran(TSV), Bonaire. After one quarter from Titan and Bonaire, Oland and Corfu were going to be on 22nm SHP.

 

[NTMBK]

Most if not all my sources can be acquired by google search engine.

http://lmgtfy.com/?q=Kaveri's+tapeout+severely+delayed
http://lmgtfy.com/?q=Krishna+and+Wichta+turbo+core

Given your track record for pulling complete fabrications out of thin air, I'm not bothering to try and substantiate your latest round of nonsense. And I like how you neglect to back up your main claim: "All projects were halted when Dirk Meyer was booted out."

 

[NostaSeronx]

Given your track record for pulling complete fabrications out of thin air, I'm not bothering to try and substantiate your latest round of nonsense. And I like how you neglect to back up your main claim: "All projects were halted when Dirk Meyer was booted out."

It is a very long story.

I'll try to shorten it..

- Blah blah blah 2006 to 2010, Board of Directors get their jimmies rustled.
- Dirk Meyer gets dropped.
- CFO takes control knows what to drop and what to delay.
- Rory Read is hired.
- CFO drops out noticing Rory Read is a snail.
- Stuff happens 2013 launches delay to 2014. More projects are cancelled.

AMD is back where they should have been in 2012-2013 with 2014-2015.

GlobalFoundries are not at fault for what AMD has gone through. 28nm SHP came on time and it came pretty early. (I do hate GlobalFoundries for calling it 28nm SHP instead of 28nm SHD super high density.)

This is my last post regarding AMD vs GlobalFoundries, etc. I had to correct mrmt since my objective is done, I am done. From this point it is only about IBM Foundry and GlobalFoundries.

 

[krumme]

If gf have usable processes. There would have been other customers. There hardly isnt. What we have witnesses is in the class of a7 designs consuming and leaking like a15. Horrible performance. Its a miracle kaveri works so well.

 

[CHADBOGA]

With various rumours swirling about Apple wanting to get in the Fab business, I'm surprised there name isn't being thrown in here.

 

[krumme]

I dont see Intels interest in buying ibm process tech. What should it be? There is nothing imo.

But who wants to be tied to gf and mubadala with a wsa? Lol. So its probably going the gf way because there needs to be consolidation i guess and mubadala will have another fancy name on their portfolio of losing investments. What a waste.

One can say the power arch can pay for expensive tech and soi and amd portfolio can use it when its more depreciated. But try to compare that to tsmc ecosystem with highend qq mobile processors ending with microowens. Lol.

If this setup of consolidated foundry tech is to make some money the ownership needs to change to some professionals like Samsung and mubadala needs to bail.

 

[NTMBK]

With various rumours swirling about Apple wanting to get in the Fab business, I'm surprised there name isn't being thrown in here.

I thought that Apple were throwing money at GlobalFoundries? http://www.timesunion.com/local/article/GlobalFoundries-to-make-Apple-chips-in-New-York-4974971.php

 

[krumme]

With various rumours swirling about Apple wanting to get in the Fab business, I'm surprised there name isn't being thrown in here.

It would be nice for Apple to use some of their gigantic mountains of cash and please their shareholders but i hope and think they want to use it a place where user innovation is in focus.

 

[CHADBOGA]

I thought that Apple were throwing money at GlobalFoundries? http://www.timesunion.com/local/article/GlobalFoundries-to-make-Apple-chips-in-New-York-4974971.php

I thought it has subsequently been confirmed that isn't happening?

 

[ShintaiDK]

With various rumours swirling about Apple wanting to get in the Fab business, I'm surprised there name isn't being thrown in here.

I cant imagine Apple wanting to do anything with fab business, besides being a foundry customer. Apple could spend endless billions on getting near to nowhere. And the money is better invested elsewhere.

 

[CHADBOGA]

I cant imagine Apple wanting to do anything with fab business, besides being a foundry customer. Apple could spend endless billions on getting near to nowhere. And the money is better invested elsewhere.

Well Apple do have the billions to spend & they might not like a scenario where Intel is on 7nm and the Android market have access to this and the best a foundry can offer is 16ff.

 

[NTMBK]

I thought it has subsequently been confirmed that isn't happening?

Not that I had heard.

 

[witeken]

Well Apple do have the billions to spend & they might not like a scenario where Intel is on 7nm and the Android market have access to this and the best a foundry can offer is 16ff.

Then they'll move on to Intel's products instead.

 

[CHADBOGA]

Then they'll move on to Intel's products instead.

Indeed, but it looks like they are doing their damndest to avoid this.

 

[AtenRa]

Today GLF's 28nm process is still inferior to TSMC's 28nm process,

I would really like you to analyze that and provide links to back it up. :whiste:

 

[Vesku]

GlobalFoundries are not at fault for what AMD has gone through. 28nm SHP came on time and it came pretty early. (I do hate GlobalFoundries for calling it 28nm SHP instead of 28nm SHD super high density.)

GF claimed they were on time but if we go by when the first 28nm GF products, RockChips ARM SoCs, were available it was at least a year late in having production level of yields/wafer. About 2 years late if you go by GF press releases ;p, but TSMC has the same silly PR version of "ready" so I just say 1 year late.

Unless the IBM execs are complete zombies they will want some combination of yield guarantees, cost/wafer ceilings, and penalties for late nodes in any WSA with GF.

 

[krumme]

Indeed, but it looks like they are doing their damndest to avoid this.

Like everybody else. How can that be?

 

[witeken]

Indeed, but it looks like they are doing their damndest to avoid this.

Why do you think that? Until recently, Intel didn't have any products to offer Apple for their iPhones and iPads.

 

[ShintaiDK]

Well Apple do have the billions to spend & they might not like a scenario where Intel is on 7nm and the Android market have access to this and the best a foundry can offer is 16ff.

Just because a company got billions doesnt mean they wanna waste it

And btw, the stockholders got a first ticket to those billions.

Indeed, but it looks like they are doing their damndest to avoid this.

PowerPC so to say will happen all over again with ARM. And they will be x86 in the future.

Also if Apple doesnt improve their marketshare, it will be very hard to justify an internal ARM development.

 

[NTMBK]

Why do you think that? Until recently, Intel didn't have any products to offer Apple for their iPhones and iPads.

We already completely derailed one thread with our arguments over Apple and Intel's future, do we really need to do that to another one?

 

[Vesku]

Just because a company got billions doesnt mean they wanna waste it

And btw, the stockholders got a first ticket to those billions.



PowerPC so to say will happen all over again with ARM. And they will be x86 in the future.

Also if Apple doesnt improve their marketshare, it will be very hard to justify an internal ARM development.

Not too hard to justify if they can keep churning out quality SoCs like the A7. Apples' current iOS strategy is sound. Control the designs and use the shear size of their order quantities to get good deals out of the contract Fabs. Of course they may regret not giving a decent bid for IBM's foundry if the contract Fab business fumbles badly leaving only Intel standing tall. Which will be the situation if non-Intel FinFet 20nm (14/16 in marketing materials) hits some major roadblocks.

 

[Homeles]

I would really like you to analyze that and provide links to back it up. :whiste:

Gate-first (GloFo) vs. gate-last (TSMC). Gate-last is overall superior.

Not sure why you're being so defensive; even GloFo wanted to go gate-last, but IBM overruled GloFo and Samsung's desires. They're finally going gate-last on 20nm, and to huge effect -- 61% lower power, and 42% higher performance vs. their 28nm SLP. The density benefits aren't worth the reduced yields and performance.

 

[AtenRa]

GF claimed they were on time but if we go by when the first 28nm GF products, RockChips ARM SoCs, were available it was at least a year late in having production level of yields/wafer. About 2 years late if you go by GF press releases ;p, but TSMC has the same silly PR version of "ready" so I just say 1 year late.

Unless the IBM execs are complete zombies they will want some combination of yield guarantees, cost/wafer ceilings, and penalties for late nodes in any WSA with GF.

Having a process ready and first product in retail doesn't go together every time. Intel manages to do it at the same time because they are both the Fab manufacturer and the retail product designer.
Also, GloFo had 32nm process and products in retail before TSMCs 28nm. GloFos 28nm is a half node of 32nm when TSMC 28nm is a full node over 40nm.
Comparing time-line of products in retail doesnt show when each process was ready.

 

[AtenRa]

Gate-first (GloFo) vs. gate-last (TSMC). Gate-last is overall superior.

Im sure you have technical data to provide ???

Edit: Not to mention that 28nm SLP (Gate First) got 40% lower power, 30% higher performance and 2x times the density of 40nm LP (Gate Last). Not only that, being a Half Node of 32nm means they can save huge amount of resources, time and money both GloFo and AMD.
Also people always forget that Gate First is 10-20% cheaper than the same process on Gate Last.

 

[Vesku]

Gate-first (GloFo) vs. gate-last (TSMC). Gate-last is overall superior.

Not sure why you're being so defensive; even GloFo wanted to go gate-last, but IBM overruled GloFo and Samsung's desires. They're finally going gate-last on 20nm, and to huge effect -- 61% lower power, and 42% higher performance vs. their 28nm SLP. The density benefits aren't worth the reduced yields and performance.

Judging by A6-5200 vs GF 5350, TSMC's 28nm seems to be a bit more efficient but it's not drastically so. Hard to get an exact comparison since the TSMC kabini is on prebuilt boards and GF is socketed but the difference appears to be relatively minor. The main impact of following IBM's lead seems to be delays, though, for whatever reason time to production doesn't seem to be near the top of IBM's priority list when planning node transitions.