There's an interesting story here on the restrictions AMD has from any further funding they may raise in the market:
AMD fund raising restricted
After their abysmal quarter, AMD's stock price shot up on the speculation of a private equity buy out rumor that's been floating around. However, a cursory look at their Q107 consolidated balance sheet shows that at current stock prices AMD is not really worth buying if the intent is to re-sell the assets unless someone thinks they can sell the IP inside the company. The other reason a private equity firm may be interested is if they think they can turn the business around. However, for a company that's already lost $1.11 in the first quarter and nowhere near any kind of profit, again this requires a huge leap of faith. Now these kind of long shots are ok when the deal is a few hundred million. But we're talking something that could end up being a close to 10 billion $ deal. Even though the guys with the money live on planet "unreal", I think there are limits to their insanity. Also, shareholders will be pissed that they are having to sell when the stock is $14...when barely a year ago it was $40.
But the real issue here is if there's no private equity deal, then AMD has to raise 2X the amount of money they actually need. And this time round the money will not be as cheap as it was when they were making profit and the stock was seeing the good times.
I have been saying since the announcement of the ATI deal that AMD would have a difficult time managing cash and debt. It feels to me that AMD management was pretty confident they would continue to make money and not have to raise further debt when they closed the deal with Morgan Stanley. The same sense of infallibility seems to be pervading how they have been guiding the market on quarterly results since the December analysts call. And it seems like they have not learnt their lesson and continue to guide flat to up for Q2. It appears they have painted themselves into a corner and if they do dilute shareholder equity through a share offering, I can see some ticked off institutional investors demanding that heads roll.
Subscribe to:
Post Comments (Atom)
8 comments:
It's going to happen. AMD BK.
"AMD also disclosed updated performance projections for its upcoming native Quad-Core AMD Opteron(TM) processors, code-named 'Barcelona.' The new Barcelona projections are based on the latest SPECcpu2006 benchmarks and show that AMD expects to have up to a 50 percent advantage in floating point performance and 20 percent in integer performance over the competition's highest-performing quad-core processor at the same frequency"
http://biz.yahoo.com/bw/070423/20070422005085.html?.v=1
Couple of problems:
1) Not sure if AMD actually demo'd this publicly or if this is more "just trust us" #'s. It does say "projections" so take this with a grain of salt.
2) "Up to 50% FP ...20% INT" Up to in my mind means absolute max and means actually average performance is likely less than these percentages.
3) At the competitor's same frequency!!! Isn't Intel going to be at a higher freq than the Barcy product that launches (aren't they already at a higher clock today?). And I'm talking current 65nm products, not even considering 3.33 (and above?) 45nm products. Is AMD just going to ignore comparisons of AVAILABLE chips because they have a higher clockspeed?
Hey my basketball team's better than your team because my best 6'3" basketball player is better than your best 6'3" player. Forget the fact that you may have better players which are taller, on a talent per inch basis I'm better!!!!
What's with all this clock for clock crap - that's useful for evaluating an architecture academically but in the real world it's your best product (factoring in price) vs your competitor's best product (and price). What if I said transmeta had equivalent performance clock for clock? would anyone care? Should anyone care?
What this information "release" means to me is that Barcy will be better on FP and most likely a tossup on INT. If AMD's top bin is 2.6GHz, Intel would need ~3.1GHz WORST CASE to match INT performance I say worst case because AMD continues with this "up to 20%" crap.
I think we now know why there were no early benchmark releases... better to wait until AFTER the quarterly financial result disaster to announce your next gen product will put you roughly on par with your competitor (and possibly fall behind in 3-6 months after it's release when Intel's 45nm parts come out) The only hope for AMD is that power is substantially better because given the native quad die size it will be difficult to compete with Intel's less costly manufacturing of their MCM chips.
I expect AMD to play up the whole clock for clock thing and conveniently distract people away from the clockspeed deltas which may eradicate and perhaps surpass any potential Barcy architecture performance improvements...
I wonder how Sharikook will react - he's ALREADY stating that the current K8 architecture is this much better than Core2 (this must mean that K10 is no improvement over K8?)
PS- sorry for the off topic post but the news is very new and I thought your readers might be interested.
180, you still fail to provide any actual analysis relating this to 2002, which AMD obviously kinda pulled out of with flying colors. What's different now? Please, finally answer this adequately in an article, instead of in your slow to update comments.
Greg - I don't understand your question...please be more explicit.
It now looks like they are taking a loan in part to pay off a loan.
Man, the more things change the more they stay the same. This sort of 'back at you' from Intel has been going on since the days of the 486.
It's all good though, without AMD - Intel would have been another Microsloth for the past two decades.
greg said: "180, you still fail to provide any actual analysis relating this to 2002, which AMD obviously kinda pulled out of with flying colors."
SIZE OF LOSS:
AMD is losing money in such as scale that it hasn't seen before compared to 2002
OVERCAPACITY:
Fab capacity planned at 25% never came to fruition.
PERFORMANCE GAP:
AMD has never lagged behind Intel by 2 or more speed bins.
MASSIVE INVENTORY:
AMD will have to consider shutting down production or writing off a large chuck of its $900M inventory
BAD OUTLOOK:
Intel's product roadmap looks promising compared to AMD. The situation will continue well until the end of 2007 as AMD can't ramp the massive barcelona.
LOAN RESTRICTION:
As 180 pointed out - adds difficulty to raise cash.
AMD never had all these problems all at the same time in 2002.
AMD offers up to $2.2 billion in notes
http://biz.yahoo.com/bw/070423/20070423006242.html?.v=1
The sky is falling!
"What's different now? Please, finally answer this adequately in an article, instead of in your slow to update comments."
Well let's see there's the massive debt (look at the amount of interest AMD had to pay on the debt in Q1).
There's the fact that 300mm is more capital intensive than 200mm. Adding additional 200mm capacity is cheaper (in the short term) then adding 300mm capacity. While 300mm is more economical - there are larger outlays of capital up front.
There's the large OEM's sucking up their capacity making it much harder to plan, increasing AMD's inventories and make it much harder to match their product mix to market demand.
And then there's their competitor who is driving their ASP's down (and instead of letting up, seems happy to keep the ASP's where they are). How much is AMD's top of the line desktop chip now?
And finally there's the restrictions on AMD's new debt (as 180 has pointed out) which limits AMD's options and effectiveness in raising new capital - this restriction did not exist back in 2002.
Finally the cost of R&D is substantially more today then it was in 2002 (I'm not referring to amount spent). AMD now has to support development across CPU, GPU and all of the various consumer electronics as opposed to 20002 where it was CPU only (though they may have been doing flash back then too?). They are also facing a competitor which is on a 2year architecture development cycle.
Post a Comment