How GLOBALFOUNDRIES thinks it will win

by Scott Bicheno on 5 May 2009, 07:00

Tags: GLOBALFOUNDRIES

Quick Link: HEXUS.net/qar3d

Add to My Vault: x

Conclusions

If GF's main objective of this introduction was to convince us it means business then it certainly succeeded. There's a start-up buzziness about the executive team that's infectious and, on paper at least, GF seems to have a very strong proposition.

But that's the big catch at this stage - Fab 2 is still just a PowerPoint presentation right now and it will be a few years before it becomes a reality.

For now it's entirely dependent on AMD for revenue and much of AMD's graphics business still goes to TSMC. It's difficult to overstate the importance of GF's first non-AMD customer. If things go well, a precedent will be set to reassure other potential customers.

If there are significant problems then companies will be even less inclined to take a risk on a relatively unknown quantity, regardless of the potential upside.

The Luther Forest site remains a mixture of pine woodland and brownfield sites today and it will be several years until Fab 2 becomes operational. In the meantime, Fab 1 in Dresden needs to be at least partially converted to operating as a foundry and must attract customers.

GF's mid-term success will depend on that but, regardless, the mere existence of another major semiconductor foundry should be good news for the technology industry and, ultimately, consumers. Increased competition means the incumbents have to raise their game. That should mean better processors at lower prices for everyone.


« PREVIOUS PAGE

HEXUS Forums :: 1 Comment

Login with Forum Account

Don't have an account? Register today!
Posted by gutieacomment - Wed 06 May 2009 02:13
This article describes what is essentially a startup trying to add competition and polarization to the semiconductor industry. There is nothing wrong with this ambition, however what it is wrong is its genesis and its intrinsic risk profile. Although AMD downloaded most of its risk onto Globalfoundries and the taxpayers of NY and Saxony, AMD's incompetent execution is rapidly showing what every experienced business person knows, that a broken business model in a market experiencing margins and profits erosion, is unworkable outside of bankruptcy proceedings. Unfortunately, unless Globalfoundries pulls soon a miracle and makes new customers materialize out of thin air, GF will reach Chapter 11 protection just a few quarters before AMD does. A sad ending for a company that once dreamed with innovating and ended up imitating to its very end.:help:

Standard & Poor's Ratings lowered Advanced Micro Devices Inc. (AMD) credit rating deep into junk territory arguing that liquidity for the company and its spinoff may not be enough to fund near-term operating losses (Saxony!) and debt requirements. “The credit agency lowered its corporate credit and senior secured ratings on the company by two notches to CCC+, in highly speculative territory” (WSJ).:surrender:

AMD has put itself in a very precarious position, mainly the company cannot benefit from its Asset Light strategy (fables operation) until Globalfoundries gains traction with new customers to offset the operating losses in the two Fabs in Saxony, mostly arising from underutilization of production capacity. That traction requires developing a new process that competes for customers with TSMC, not only Intel. Moreover, there is a huge pricing conflict between AMD and Globalfoundries for MPU wafers, a situation that may not get resolved in time to save both companies. If AMD continues to support Globalfoundries using above market prices for processed wafers, it cannot become viable in light of Intel’s cost advantage and aggressive pricing, and thus Ch 11 followed by liquidation is clearly on the horizon. Instead, if AMD decides to sink Globalfoundries, it will have to renegotiate with Intel its licensing agreement to allow production at TSMC or Chartered. This is not very likely, and at the rate both companies GF and AMD are crashing, since both of them will be out of cash by 2010-11 and 2011-12 respectively.:juggle:

I suspect that Globalfoundries will have to be sacrificed at some point, thus allowing AMD to migrate production to TSMC or Chartered, provided that Intel allows it; which is in itself very unlikely.

The situation of both companies seems to be a puzzle without a solution, other than having IBM spun put Microelectronics very soon, merge with Globalfoundries and provide a combination of corporate and Emirates financing to the combined company. This is a hugely speculative scenario.:eek:

Either way, I suspect AMD will have to abandon competing against Intel by 2011 or 2012. AMD investors are now all aboard an imminent train wreck, and stand to be wiped out by 2011 at the latest.:angst:

The questions that taxpayers in NY and to a lesser extent in EU counties should be asking is: Why did NYS accept the conditions of the revised deal offered in Oct. 2008? Why the overseeing authorities did not end in Oct 08- Dec 08 a deal that was so clearly skewed against the interests of NYS taxpayers? Why was the modified deal approved without public discussion?

Also , Why did the EU poured huge subsidies for so long in a company unable to secure leadership and protect margins in the MPU space? What has it been the outcome of similar subsidies for Quimonda, Infineon and now AMD?:censored:

Dr. Milton Friedman would say that these misguided corporate subsidies are just another poster case of why markets and not governments should pick winners.:surprised: