By: David Kanter (dkanter.delete@this.realworldtech.com), January 19, 2011 11:53 am
Room: Moderated Discussions
Azazel (none@gmail.com) on 1/19/11 wrote:
---------------------------
>David Kanter (dkanter@realworldtech.com) on 1/18/11 wrote:
>---------------------------
>>Almost all the stuff you listed are implementation details which a CEO may or may
>>not have a great deal of visibility into and control over.
>>
>>The spin-out of GF was a pretty big deal and was handled reasonably well. I think
>>designing a mobile oriented microarchitecture was a very good idea, and something
>>long overdue at AMD. So Dirk deserves credit for backing that project as well,
>although the jury is out on the results.
>>
>>Also, while some of AMD's products are somewhat late - that's not exactly new.
>>The K8 came to market late and did quite well. Bobcat derivatives seem to be on
>>time, while most of the problems seem to revolve around the gate first 32nm process.
>>IBM and GF's 32nm process choices weren't great, but it's unclear how much influence
>>AMD had on those decisions relative to IBM.
>>
>>So I think you're being unduly negative and pessimistic. But the last 5 years
>>certainly weren't a walk in the park for AMD either.
>>
>>David
>>
>I'm neither pessimistic nor optimistic. I'm realistic.
>I see nothing positive in AMD's development for the last >5-6 years. The absence
>of accomplishment examples only proves that hard to find >even one.
They now have a reasonable product for mobile/tablets/netbooks.
They don't have fabs any more.
They have an integrated CPU/GPU.
I think AMD is in a stronger position for client systems than they were in 2006.
>As to GloFo, that deal was of Ruiz's responsibilities. In >my view, the deal was done in the very bad way.
>AMD had had no intention to sell its fabs back in 2006 >when she bought ATI.
I don't see anything wrong with the GF deal. It's true AMD didn't get a lot of cash, but they got some cash and they also got the buyer to commit to spending a ton of cash further down the road on the manufacturing. ATIC basically made a $5B+ commitment (I don't recall the exact figure), which is pretty nice.
>That's
>why GloFo deal was not a part of AMD's strategy at that >time. AMD was compelled
>to sell their fabs due to the burden of huge debts.
That was definitely involved. But that doesn't make it a bad idea. If you look at the trend, AMD is hardly the only company getting out of their manufacturing. How much $ did Fujitsu or TI get for their digital fabs?
>And because that was done two
>years later, when AMD shares experienced a sharp decline, >AMD had gotten only about
>2B$ for ~67% sold stake. It's a joke of money for the >modern fabs. BTW, now AMD has 14% only.
A fab is only modern for a few years...especially a CPU fab.
David
---------------------------
>David Kanter (dkanter@realworldtech.com) on 1/18/11 wrote:
>---------------------------
>>Almost all the stuff you listed are implementation details which a CEO may or may
>>not have a great deal of visibility into and control over.
>>
>>The spin-out of GF was a pretty big deal and was handled reasonably well. I think
>>designing a mobile oriented microarchitecture was a very good idea, and something
>>long overdue at AMD. So Dirk deserves credit for backing that project as well,
>although the jury is out on the results.
>>
>>Also, while some of AMD's products are somewhat late - that's not exactly new.
>>The K8 came to market late and did quite well. Bobcat derivatives seem to be on
>>time, while most of the problems seem to revolve around the gate first 32nm process.
>>IBM and GF's 32nm process choices weren't great, but it's unclear how much influence
>>AMD had on those decisions relative to IBM.
>>
>>So I think you're being unduly negative and pessimistic. But the last 5 years
>>certainly weren't a walk in the park for AMD either.
>>
>>David
>>
>I'm neither pessimistic nor optimistic. I'm realistic.
>I see nothing positive in AMD's development for the last >5-6 years. The absence
>of accomplishment examples only proves that hard to find >even one.
They now have a reasonable product for mobile/tablets/netbooks.
They don't have fabs any more.
They have an integrated CPU/GPU.
I think AMD is in a stronger position for client systems than they were in 2006.
>As to GloFo, that deal was of Ruiz's responsibilities. In >my view, the deal was done in the very bad way.
>AMD had had no intention to sell its fabs back in 2006 >when she bought ATI.
I don't see anything wrong with the GF deal. It's true AMD didn't get a lot of cash, but they got some cash and they also got the buyer to commit to spending a ton of cash further down the road on the manufacturing. ATIC basically made a $5B+ commitment (I don't recall the exact figure), which is pretty nice.
>That's
>why GloFo deal was not a part of AMD's strategy at that >time. AMD was compelled
>to sell their fabs due to the burden of huge debts.
That was definitely involved. But that doesn't make it a bad idea. If you look at the trend, AMD is hardly the only company getting out of their manufacturing. How much $ did Fujitsu or TI get for their digital fabs?
>And because that was done two
>years later, when AMD shares experienced a sharp decline, >AMD had gotten only about
>2B$ for ~67% sold stake. It's a joke of money for the >modern fabs. BTW, now AMD has 14% only.
A fab is only modern for a few years...especially a CPU fab.
David
