[Beyond3D Article] AMD Q1 2007 Warning Analysis

Much worse than I expected. The big question for me remains...INVENTORY. If it is really high they are in a load of trouble. After hours trading is only down like $.30 cents which is puzzling. I would have expected $10-$12.

 

Considering AMD's prior excuse for bad performance was unexpectedly poor channel sales, and the news of additional loss in marketshare, I'm not getting a good feeling about inventory.

 

28% (31% excluding charges)? Are you effing kidding me?

 

Inventory is $937 million up almost 25% from Q4 2006. 40% of this inventory is 65 nm. Their intention is to work off this inventory throughout the year by cutting production.

They are also going to take the entire $500 million reduction in CapEx in the conversion of Fab 30 to Fab 38. In addition headcount will be reduced as they restructure.

 

If they're cutting production, that sounds like fab underutilization's going to bite them in the ass.

Underutilization would also mean that none of the new cores are near volume production to pick up the slack.
One would think Barcelona or Agena with the larger die size could be produced in greater amounts to sop up some of the wafer space, but cutting production indicates that this is not the case.

 

Barcelona 2 core and 4 core desktop parts out in Q3. New mobile part out in Q2 2008.

R600 launch will be hard with product available. "We don't do soft launches". 2900 XT and down was the implication that I got with a nice barbed dig at Nvidia concerning drivers.

 

I wonder how much R&D money graphics division will get allocated to it going forward.

 

BTW, Arun:

Actual results: Computing Solutions: 918; Graphics + Consumer Electronics = 315. Which brings up the question: did the chipset business get rolled into computing solutions? I have hard time believing that entire ex-ATI has shrunk to a $315 million operation.

 

Richards said on ATI graphics

"Units was up and ASP was slightly down. Frankly, it is logical since we are waiting for the launch of our brand new R600 line of products, which will enable us to restore the price positioning in the marketplace. "

"Restore the price positioning" is management speak for " bunging the price right up "

I think that if R600 is a good product then that will be a nice foil to K10 if that also is a good product. Two great products should regain perceptions of the company in the eyes of buyers.

What worries me though is that K10 will probably be around for quite a while and with Intel doing a new architecture every 2 years and the next one being in 2008 it maybe a long time with Intel having the upperhand after a short reign by Barcelona. That's assuming Barcelona is somewhat faster than the tweaked 45nm Intel Core products....

AMD really need to make hay while the sun shines, they are competing against two very aggressive competitors though, no easy task.

 

I didnt hear the cc, but cnbc mentioned some talk over AMD allowing some private equity stake. Might be why the stock isnt taking a hit right now.

epic

 

QFT, its would be crazy to buy shares in this company at this valuation. Its too wild a ride for me.

 

In server at the 4-8 (maybe a little for 16) socket market, cores related to Barcelona should perform better than current Core2 chips, and could possibly be competitive against the early 45nm derivatives. This will be workload dependent, it is likely that Kentsfied and Cloverton will still beat a Barcelona system at a number of workloads.

2-socket is more likely favoring Intel in some workloads at 65nm, probably totally in hand by 45nm, especially with some upcoming chipsets.

The situation doesn't seem likely to change for the better with Penryn.
With expectations that Barcelona will not make a huge impact in 2007, it is likely that Barcelona's critical period is going to overlap with an aggressive ramp of cooler, faster, and cheaper 45nm Intel chips.

The Barcelona line was the one that would have been good about 2 quarters ago against Conroe's ilk, not Penryn's a half a year from now.

For the desktop, Agena or Kuma is not likely to beat Penryn, and unlikely to beat Conroe. Barcelona's big changes deal with FP and SSE processing and quad-core optimizations, but the integer side is not radically different from K8.

Integer single-threaded performance is likely better than K8, but there is little indication and no definitive statements from AMD that it will be better than Conroe, and a number of hints and ass-covering vague statements to indicate that it will not be.

AMD's all about the platform and scalability, none of which will affect the desktop for quite some time.

 

"Restucturing" is somethign I read as "staff cuts".

 

Who can they fire, or what can they sell off?

The former ATI chipset business? It's not too great right now, and it would likely fetch a better price if the purchaser was on Intel's good side.
Just getting out of AMD's control would open 80% of the chipset business that was closed by the buyout.

The small mobile products group is in a growing field. It's not enough to buy AMD any revenue on its own, but it could fetch a decent price from someone who in a less treacherous situation.


Who can be made redundant?
Can someone characterize how much AMD was able to cut from ATI? I've heard mention that it wasn't as many as was hoped.

What hope does R700 have for timely development when AMD is trying to survive as a CPU maker?

 

One is question is would there be a buyer fo ts Dresden fab. Chartered and IBM come to mind as possible suiters, as well as perhaps TSMC, but prhaps there are resricitons from AMD's partnerships. Other than that there does not seem to be much to sell.

However, if the could sell the fab, that would allow them to pay off a bg chunk of debt and get rid of a lot of operating and fab-related r&d expenses. If the is viable, the sooner they can do this the better, because they are otherwise looking at a Chapter 11 scenario, because as Arun points out, it would be crazy for private equity to buy at these valuations and they are just bleeding money all over the place. The will not have money to fund operations in fairly short order.

The decision not to go fabless yeras ago when Hector took over from Jerry does not look bright. Total lack of vision or guts. The escalating costs of fabs has not been a secret.

 

They aren't going to sell the chipset business, that'd be crazy for them and invalidate a major part of why they bought it in the first place.

I suppose they could sell DTV and handheld, but that'd ruin some major synergies as well and would show signs of panic.

They're talking about on the order of 500 employees attrition in the nearish term, which is fairly minor in the large scale.

It seems very clear that they don't see a reason to panic at this point, or the actions we'd be seeing from them would be much more serious.

 

Just a bit of crazy talk, but i wonder if IBM would be interested in acquiring AMD. There might be some sharing of patents/ideas between GPUs and their supercomputer chips.

 

If AMD decides to cut capacity, it may as well give up on x86 altogether. Intel's triple threat of more fabs, better processes, and better chips will go unanswered.

AMD will not be able to get the volumes to amortize its design costs, and without a leading-edge fab, its chips will always underperform.

See VIA for an example, and see Intel's planned very low power chips that will draw less power, outperform per clock, and clock higher than any current VIA chip. At the volumes they'll sell, it's also unlikely VIA will compete all that well on price, if Intel decides to make a bargain line.

AMD legally can't go fabless and continue to make x86 chips, per its licensing agreement with Intel.

There is no future for AMD's x86 if it goes fabless. If that happens, it'll just be ATI with a whole lot more debt and a bunch of employees it will soon fire.