blauboad -

Exactly how do you know this? Are you stealing my mail?

No. I'm not competing with you. I'm just trying to make money.

BTW, AMD is going down this morning. I wouldn't be surprised if a disappointment is in store, at least compared to what has already been factored into the price. More Puts to be written tomorrow...

blauboad

I accept this as your concession of defeat.

What were we fighting about? I'm trying to make money and it looks like I'm winning.

blauboad -

Another good reason to suspect that 90nm is in trouble.

You guys don't need good reasons. It said so in the inquirer!

blauboad -

unless Prescott is embarassingly slow now and they can't make it sellable without exceeding the already heightened power spec. Then your scenario would work :)

Then Intel would have done a 90nm shrink of NW and enjoyed a nice ~10% performance boost.

Jerry -

I have a strong feeling AMD will "promise" breakeven in Q1 next year and profitability in Q2 and will guide as such in Thursday's CC.

Yes, just like they promised last year, and the faithful will be jumping for joy, just like last year.

Jerry R -

Did you hear this in the CC?

Yes

Sgolds -

we got the flash business that Intel missed, they got the microprocessor business we missed.

Pretty hard to claim you are objective when you talk of "they" and "we". Otherwise I agree with your assessment.

8-/

The real 'bang' will come from forecasts, from the CC, of Q1 and Q2 '04

Oh please!

The sad thing is that there are people that will actually believe the promises...

Tenchu -

Too bad Banias didn't "pull it's weight"...

YB -

I'm thinking about writing some covered calls tomorrow. What do you think?

If Intel gained market share like they think they have, AMD may not impress so much tomorrow. That means CCs may be a good idea.

Intel still on schedule to ship for revenue both desktop and mobile products on 90nm this quarter.

I read this to mean that Prescott is not delayed to Q1/2 as claimed by tabloid websites.

Keith -

Now, isn´t that great? INTEL raises their flash prices at a time of oversupply to lose a chunk of marketshare, and now that a flash shortage seems to be developing and prices are said to finally stabilize, they start dumping inventory into the market and start a price war, with the logical consequences.

Let's see, how many unsubstantiated rumors can you start with one post?

#1 Flash shortage
#2 Stabilizing prices
#3 dumping inventory
#4 Intel starts price war

You must have gotten this from theinquirer, right?

InternetPlay -

Sorry that I don't have much time to spend on this board, so I don't know if this one has been posted before -

Yes, it was posted about a year ago when it was written. Did you look at the date?


posted 11:33am EST Mon Sep 30 2002

U.S. companies with subsidiaries abroad lobby hard to cut a tax on foreign profits that they say keeps them from bringing money home .....

Big sums are at stake. Intel has accumulated some $6.3 billion in untaxed profits earned by foreign subsidiaries, according to its most recent annual report to the Securities and Exchange Commission.

http://www.oregonlive.com/business/oregonian/index.ssf?/base/business/1065873918272280.xml


Does anyone know if these funds are included in Intel's reported cash position?

Chipguy -

If I remember correctly, N10 development was going on concurrently with the i486. In fact both shared the same building and floor, SC4/2FL. The N10 was introduced at a major forum and billed as the world's first 1 million transistor device. The i860 enjoyed quite a bit of success from it's FP capability and some neat 3D instructions. There was a i860XL that followed but the product line died thereafter. As I remember it was viewed as a speculative experiment, kind of like the i432. As long as we're waltzing down memory lane, let us not forget the origins of the i960 from the days of the Biin project. The i960 was a fine processor if I remember correctly and the first to go superscalar. When Biin died Intel marketed it as a uC rather than compete with the cash cow. So at one point Intel had 3 architectures simultaneously.

Petz -

Please come back when you have a better source than tabloids.

wbmw -

looks like you copied and pasted the wrong quote.

Yikes! Sorry.

Sgolds -

Now, Elmer may believe that is a year from now, but everyone else is betting that is 6 months away.

They're the same people who believed Hammer was coming out in late '01. I know you read that PDF someone linked to yesterday. It had AMD converted to almost 100% 90nm wafer starts by Q4'03. Do you really think AMD believed that when they presented it or was it intended to pull the wool over the eyes of the analysts? It's baloney like this and countless others that makes people skeptical of anything AMD claims. I've been watching it for years and I expect you have too but you believe "this time it's different". I believe it's more of the same.

Joseph -

How do you work down inventory (without resorting to the dumpster)? Less wafer starts.

Chipguy or WBMW could do a far better job than I.

Sgolds -

How do you work down inventory (without resorting to the dumpster)? Less wafer starts.

Yes you're right and I've tried to include that as an alternative explanation. AMD volumes into a demand driven market are the real measure of their capability.

Paul -

Knowing Intel's methods, if they didn't think Banias could be manufactured to it's target spec with predicted yields they wouldn't have done it. It's a standardized process that is finely tuned to yield. Same process as P4, same design rules, same yields (read defect density). I don't comment on what Intel's yields really are but they have more money than anyone else to throw at the problem. One might expect that they are getting something in return. The perception in the industry is that nobody else is even close. I will tell a story though that I think I can share after a few years have passed. Years ago I was involved in some engineering issues involving a chipset produced by Intel (incidently this is the highest volume product ever shipped on a single stepping in the history of the industry). I had the ability to track the yields. I was simply astonished to see that it was running at a defect density of only about 1/2 what was considered by others to be World Class. That was a few years ago and I'm not saying Intel is running at that level today, but again I'm not saying they aren't, I'm just not saying.

Paul -

In the absence of some custom circuitry, a defect doesn't know if it's fallen onto a P4 or a Banias. They don't discriminate. They're an equal opportunity part killer. It's defect density driven and not product specific. If Intel can make 30-40 million P4s a quarter they can make 60-80 million Banias'. It's purely a business decision how they allocate their capacity.

Chipguy -

Production limited at 2m for a 85 mm2 device made on 300 mm
wafers when Intel is cranking out tens of millions of P4s per
quarter?

Banias is not on 300mm wafers. I don't know why but it isn't. Nevertheless if Intel has anywhere's near good yield then 2M per quarter is childs play.

Sgolds -

You better hope the wafer start number is wrong because the low volumes of XPs would be very poor on bulk silicon.

Alan -

The P4EE is a fast Xeon but it also supports an 800MHz FSB

Petz -

Even Intel admits that they oversold the Centrino.

I think you misunderstood a quote. I think they said they oversold wireless.

Sgolds -

It isn't that Intel can't make more of them, it just that it is not their product direction. If 90nm were on schedule and performing to spec then there would be no EE. IMHO.

Well you may turn out to be right, time will tell. I see nothing wrong with adapting. Another reason they may be doing it is because they can.

Sgolds -

Normally, Intel outsells AMD 5:1. You are suggesting a 1:1 ratio of EE to FX. This shows it is not a serious product by Intel.

That would be tue if it was intended for a broad market but the highend games market is quite small.

Chipguy -

I agree with you in that I see no pressure for Intel to move to 90nm at this time. If your view of Intel's yields and manufacturing costs is correct then there is no immediate reason to push something out that may or may not be ready. AMD is unable to put pressure on Intel until 90nm and in my opinion they don't seem to have 130nm SOI worked out yet so 90nm is probably quite a ways out despite the announced schedule.

Sgolds -

at that price they certainly won't have to worry about selling many EEs! Just more evidence that it isn't a real product, just a dog & pony show.

You are easily convinced by your own conclusions. As you assuming Intel has priced it high because they can't make many of them? I wouldn't be so sure if I were you.

Sgolds -

I don't know if 50% of Dresden is typically in such transition stages, but I would not be surprised given how aggressively AMD converts to a new geometry. They are in the middle of their 90nm conversion now, with first retail product coming in Q1 (rumored) or Q2 (promised)

You may believe that but you're a rare breed.

Jules2 -

This time I got in early at 4.42. I'm going to stay long.

Good for you! I wish I had too.

Petz -

You may be right and you might make a ton of money. How about posting your positions? I don't mean actual shares but are you long and at what price did you get in? You can surely find some instances of where AMD has had some temporary successes but the facts speak for themselves. AMD is a net loser over the last 30 years and the share price has gone essentially nowhere over that time period. Intel makes more money in 1ps than AMD has made in it's entire history. While the past is no guarantee of future results, and the race doesn't always go to the fastest horse, it's the way to bet. A horse that loses race after race could possibly win the next one but only a fool ignores it's track record. You place your bet as you think best and I'll do the same.

borusa -

So are you saying AMD is a speculative investment?

You and I may have different investment objectives. I don't need to speculate because I'm not trying to get rich anymore. Instead I think AMD is a fine cash generator. I don't expect AMD to ever be terribly successful so I expect it to continue trading in this general range. While I could eventually turn out to be wrong, I've done quite well writing Puts and Calls as it stays in this general range. If you're in the speculation mode then I wish you luck.

borusa -

I think the point here is that if anyone has followed AMD for any length of time they know the empty promises and repeating disappointments. As you probably know the company's earnings (ha!) release next week will almost certainly make AMD a net loser over it's entire corporate history. There is a pattern there and don't shot the messenger for pointing it out. Only the gullible would believe the next promises of profits that are sure to come next week. Did you forget that last Q2 was supposed to be profitable?

When AMD releases their losses next week I expect a moderate pullback. I will be looking to sell $10 puts at that time. Why if I'm so disbelieving? Because I've been watching AMD long enough to know that they'll promise profits are coming and there are plenty of people who will believe them. Enough to hold up the share price.

Jerry R -

I guess the real question is whether is will break through $30.00 and hold, which I'm sure is what all of us Intel longs here hope.

Well almost all longs. I have $30 CCs out there. If you don't mind I'd just as soon it waits until after next week...

Jerry R -

On a related note, the consensus estimate for AMD seems to -0.37. If this is true, then this will have been the best quarter since Q3, 2001 (-0.14). I will be very curious to see what Q4 guidance they provide next week, and when they predict profitability.

Before I comment I'll just add that I am an AMD investor from several angles and I rely on AMD going nowhere to make my money, as such I have zero confidence in any forward looking claims AMD makes. I've been listening to that baloney for too long. How many times have we heard that profits are just around the corner and the yields are fine? Anyone who doesn't relate to that just hasn't been listening for the last decade or more. While anything is possible, and success is not impossible, AMD has been promising so much for so long while delivering disappointment after disappointment that anyone who believes their promises is just a sucker who gets what they deserve. IMO, there is money to be made in realizing the pattern is likely to continue. AMD keeps promising profits and there is always a new crop of people willing to believe them. The current crop gets disillusioned and moves on and there is a new group of enthusiasts ready to start the cycle all over again. Profit from them. There is money to be made here.

Petz -

BTW, Mike Magee confirmed on SI that the CPU quantity estimates that he reported were a communication from AMD to Asian motherboard makers.

But the question is according to whom? Did AMD confirm it was a "communication from AMD to Asian motherboard makers" or did the board makers confirm?

Sounds like just more tabloid rumors to me. That means it will be taken as fact here.

Alan -

I think 5500 WS/w is also reasonable with an assumption of something like 200WS/w going to 90nm development, and a drop in available starts due to the increased process complexity, offset by the expansion... Does it sound like soup yet?

This is much closer to what I would estimate too. The joke is that AMD apologists have estimated up to 50% of Dresden's capacity going to process development and process conversion on an ongoing basis. This has been necessary because terrible yield has been the only other reasonable explanation for their very poor output assuming there is demand.

Sgolds -

We've discussed this - when a company has a single megafactory that covers introducing new product, mass production and conversions to new processes then you can't simply multiply floor space by some constant to get maximum yield.

We all understand this. The problem is that for many this mysterious variable is allowed to rise and fall as necessary to explain away the missing output while never allowing yield to fall in the equation. This missing capacity is always used to explain the descrepency, no matter how large it is. Sometimes it is well over 50% and that simply doesn't wash. AMD is known to have expanded their floorspace yet the output never seems to get above the 50% point.