Semi, the only piece of the analysis that was left out (and its a big piece), is the increase in the number of chips produced per wafer with the smaller geometry process. This presumably represents a great deal of incremental revenue.
Question is will the extra revenue offset all the headaches you outlined? Is it in fact like buying a new fab on the cheap?


I'm no Industrial Engineer, they figure out this Layout/Throughput/Labor Investment return, so it's a WAG, but I think the investment return may still not be there. The reason I say that, is that while smaller geometry process would generate more chips per wafer, the scenerio of using equipment that previously manufactured microprocessors, to manufacture Flash, would return less revenue, since Flash sells for less than microprocessors. You'd have to crank out allot more flash chips to make up the revenue difference.

If you were reusing the equipment to make cheaper microprocessors, then the investment return might be worth it. On the other hand, the equipment itself would likely already be fully depreciated and paid for, and if the people Installing and Qualifying the equipment were already familiar with it, to minimize the issues...... it might pay off. It would be close.

So, my bottom line WAG is, as a Bean counter, in a company with tight resources, and not much money for new equipment, I MIGHT go for it. As an Install/Qual Engineer, or as a Process Engineer, I'd hate it. So if I were AMD, in their current financial position, and it were my decision, I'd get my best Equipment and Process people and try it....... One time. If it worked, great, if it didn't, Not much to lose, eh?

Semi

What happens to the old tooling at Dresden (i.e. the 180 nanometer stuff and now the 130 nanometer stuff)? Can't it be shipped to another plant and used to produce flash?

Maybe. At the beginning of a new process, it might be possible, but once a process is up and running, I think it would be difficult to incorporate a different piece of equipment. Since AMD does not follow a Copy Exactly Philosophy, the last I read, AMD has different equipment at each Fab. Transferring the equipment, would require de-installing the equipment, shipping it to the new site, installing it, and qualifying the equipment to the Equipment manufacturers specs. Then it would require modifying the current flash process to work on the at least 2 different tools at whatever manufacturing step it is being used on, perhaps several steps.

This is a Process Engineers nightmare. Have different pieces of equipment running the same process step in a Fab creates difficulties, especially when things go wrong. Is it the Process? The tool? Will containment and solutions work on both tools? How would you know that a "fix" on one tool won't crash the other?

Improvements to the Process could also have to be done twice, once for each toolset, increasing labor costs, and delaying improvements Will the improvements work on both tools? Even if they work, will the End Of Line results be equal? What do you do if they're not? Modify the Process for each tool?

Alternately, You could use the equipment to take over all the Processing of a particular step, and not share that step between 2 different tools, but again, that would require requalification of the Process at that step, on a new toolset. Are the yields going to be the same? How long is it going to take to requalify the Process? Will the tool reuse cost savings be enough to offset the cost of requalifying the process?

Some equipment can be reused from Process to process. Recently, the problem has not been reusing equipment from Process to Process, as much as 200mm to 300mm. Based on all the issues I've highlighted above, I believe that it would probably be more cost effective to sell the equipment for whatever you can get, and be done with it.

Semi

Centrino resulted in less than a 1% increase in market share, even using the quarter over quarter numbers.
Petz


Not interested in any further discussion of Centrino Market Share with you.

http://www.investorshub.com/boards/read_msg.asp?message_id=1339532

Semi

Looking for the next "clue" -- its right in front of your face.

It's hopeless...... Fine....... I give up....... Forget it. You're right, Centrino isn't selling well, it's not pulling it's weight, there was no sequential market share increase for intel due to Centrino in Q2, IDC is wrong, It's all a farce.

And there really is a Santa Claus.

There..... Happy now?

Semi

If sales go up then INTC has to be able to produce to meet demand. The item they questioned was availability not demand. I didn't say it but I am asking that those that reply on the subject at least keep on the point of discussion and not skew the discussion with facts that aren't relevant.

Just don't fly off the handle and cram market data down someone's throat when they voice an idle thought. Maybe just maybe even an erroneous premise could lead to a valid discussion of fact if you don't take it as a personal threat to one side or the other.


The fact that one vendor does not have a Centrino Notebook greater than a certain MHz says nothing about availability of the product. Maybe the vendor didn't anticipate the surge, and didn't order sufficient notebooks from the supplier, maybe the notebook supplier didn't anticipate the orders, and didn't put in the orders to the Motherboard maker. Maybe the Motherboard Maker didn't anticipate the Notebook makers order and didn't make enough boards. Lots of things could have happened besides intel not being able to supply the chips.

As far as your contention about Petz's question, Maybe you're right...... Perhaps Petz was just "voicing an idle thought"........ Perhaps Petz's past history of criticizing Centrino was not the root cause of his comments........ Perhaps Petz's subsequent post on "Centrino Lackluster Demand" was not intended to suggest that "Centrino Is Not Pulling It's Weight". Perhaps it was all just an idle thought. Perhaps there really is a Santa Claus. But I doubt it.



Semi

And what is the "hard data" that chipguy is referring to? You apparently didn't read the following:

Apparently, it is you who didn't read it too closely. If you notice, that data the inquirer quotes is "In the Year Ago Quarter", and IDC was "Sequentially" from Quarter to Quarter. ALL CPU sales fell Year On Year, so compared to a Year ago, when Centrino wasn't around, yeah, Centrino didn't help much in a depressed market. Quarter on Quarter tells a very different story, doesn't it.

Semi

CJ, CJ, CJ..... Right..... It's all part of a great conspiracy by The Evil Intel, simply to make it look like with Centrino, intel is doing something right. You forgot a few rules for this fantasy though.....

Wait, I forgot one more.....

A price cut for AMD means a new Speed Grade Introduction, or Lower Manufacturing Costs, or Improved Yields, but for intel it means "Lackluster Demand"

Semi

More evidence of lackluster Centrino demand: Intel to cut Pentium M prices in October
by Geoff Gasior - 04:21 am, August 19, 2003

Or...... The 0.13u Cost cutting, which has been going on for the last quarter, has shown enough cost savings to allow intel to cut prices on chips that are starting in the Fabs right now, and sell more chips on the run up to Christmas.

Or..... Yields have improved so much, that intel can now afford to promise to fill orders for October and sell each each of those chips for less.

Or..... Intel is readying a speed grade increase in October, and is cutting prices prior to the introduction of the new speed grade.

See Petz, there's lots of other possible reasons..... Besides yours.

Semi

The data you hold onto so dearly is confirmation of what happened in the past. The clues the people on this board are searching for are indications of current and/or future performance. I might trust a market share report but I don't trust market share projections and thus I keep reading looking for that next clue...

That data was for Q2. Market research data for Q3 has not been posted, but "The next clue" is there, IF a person cares to look..... Objectively, that is.

It has been widely reported that Computer sales are on the rebound. Now I'm sure that SOME may believe (or more accurately.... Hope) that all of a sudden Centrino is going to do an about face, and go into the crapper, but based on reported turnaround by motherboard makers, and reported improvement in overall sales, and estimates of Mobile gains in general vs. desktops, and the spread of WiFi to even more locations, that doesn't make sense, now does it. Clue enough for AMD Fans? I doubt it.

And why do I have a feeling, if the situation was reversed, and it was AMD mobile gains being reported, then none of you would have an issue accepting, and cheering for, the "Data that I hold onto so dearly" Hummmmmm?

Semi

No, according to Best Buy's website NOT A SINGLE BEST BUY STORE has a Centrino notebook faster than 1.4 GHz. We're not talking about a single store here.
Petz


SO WHAT! Obviously there is no interest here in accepting a basic FACT.....

That according to IDC's Market Data Research, Centrino is selling well, and helped intel take mobile market share from AMD.

THEREFORE, YOUR PREVIOUS CONTENTION THAT CENTRINO IS NOT PULLING IT'S WEIGHT IS FALSE.

Geeeeezzz

Semi

Petz, Petz, Petz. You know the rules. If you can't find Intel stuff in the stores, it's because demand outstrips supply. If it's AMD, it's because there isn't any demand and/or they can't make them.


CJ, CJ, CJ..... Right..... It's all part of a great conspiracy by The Evil Intel, simply to make it look like with Centrino, intel is doing something right. You forgot a few rules for this fantasy though.....

All market data that shows intel gaining share is wrong, or paid for by intel.

All market data that shows AMD gaining share is Gospel.

Everything Van's Hardware says is unbiased.

All AMD Motherboard supply issues are the result of intel strong arming.

Intel's compiler was specifically written to cripple AMD chips.

All benchmarks are corrupted by intel. Unless AMD wins, then it's Gospel.

All vendors are only selling intel because of rebates.

Dell locks out AMD only because intel is giving the chips away.

Jerry Sanders Deserves all those Stock Options

None of Intel's management deserves any options.

Semi

Oh yeah, I forgot..... AMD ROCKS!....... GO AMD!

Why doesn't the largest electronics store in the US sell any Centrino laptops in its stores faster than 1.4 GHz?
Petz

I don't run that store, ask them.... Besides, what was being discussed was overall Centrino sales, and as long as Centrino is IMO certainly pulling it's weight, by taking market share from AMD, according to Market DATA, that's what really should matter to Intel and AMD investors, not some transparent attempt to deflect that fact, with some B.S. about why a speed grade, is or is not sold at a particular vendor.

Semi

I'm not sure exactly what you mean by "pulling it's weight". Since Centrino is new I'd expect it to be "adding some gravy". OTOH, more time is needed. I hope it's not "puking it's guts".


Those are Petz words, not mine, ask him. I guess he can make it mean anything he wants to twist it into, but the fact remains that according to published DATA, Centrino is selling AT LEAST as well as intel expected at this point, and probably BETTER than expected. No matter how "someone" wants to twist it, by any NORMAL standard, that qualifies as "Pulling it's weight". And as far as "Puking it's guts"...... Not likely. With further upgrades to Centrino Wireless protocols in the pipeline, it is MORE likely, that Centrino will continue to take market share away from it's competitors. The right product, at the right time.

http://www.investorshub.com/boards/read_msg.asp?message_id=1338863

Semi

I still contend the Centrino is a tough sell becasue of it's low MHz.

--- The Data does not support your "Contention"
http://www.investorshub.com/boards/read_msg.asp?message_id=1338863

Semi

Petz said Centrino is 'not pulling its weight' -- because its TRUE.



---IDC Research, an UNBIASED MARKET RESEARCH source, does not agree with YOUR assesment.


http://story.news.yahoo.com/news?tmpl=story&u=/nf/20030818/bs_nf/22106
Centrino Success
The IDC report demonstrates that Intel's Centrino release, buoyed by a massive publicity blitz, was a plus for the company.

"Intel's ramp up of the Pentium M processor, part of its Centrino platform, showed in a sequential increase in mobile-processor shipments," said Shane Rau, IDC senior research analyst.

"Meanwhile, AMD's mobile shipments reflected the seasonal decline in demand and loss of share by falling off approximately 25 percent sequentially. As a result, Intel took substantial mobile-processor share from AMD," Rau said.

The report indicates that Intel increased market share in the mobile-PC processor segment by nearly 3 percent, to 88.1 percent. AMD's current mobile-processor share is 10.6 percent.



---So, in my opinion, if people have a choice of believing what Market Data says, or what You say, I think most people will go with the Data. Except Bonefish, of course.

Semi

Nothing, necessarily, but the symptoms of the problems point to keyboard firmware and apparent bugs inn the touchpad
I could be wrong, but that sounds like chipset, MOBO, or driver design. Chipset is Intel 855PM. Going to be interesting to see how this turns out.
INTC's had some chipset problems in the last year or so IIRC.
When I look at the specs of some other Centreno notebooks it looks like they mostly have the 855PM chipset. Are other chipsets available for the Centrino or is it a package? Who would have designed the drivers?
8-/


Not according to this link. Keyboard firmware is not controlled by the MOBO chipset, but by a keyboard controller chip, internal to the keyboard itself.

http://www.pctechguide.com/14input.htm
The keys are connected up as a matrix, and their row and column signals feed into the keyboard's own microcontroller chip. This is mounted on a circuit board inside the keyboard, and interprets the signals with its built-in firmware program

Semi

I was going to jump on it but Keith beat me to it.
Whatever the problem is with the x1000 it's bios/mb/periprials not the centrino that's to blame.


Thank You.......



Semi

No Centrinos in sight.
Went to Best Buy yesterday. 12 Notebooks...no Centrinos.
Went to Costco today. 5 Notebooks, no Centrinos.
Where are the Centrinos Intel keeps telling us are selling so well?
Bone


SIGH....... Do we need to go through this whole "Centrino is not pulling it's weight" thing AGAIN? I recall you chirped right in claiming Centrino would be a "weak seller", when Petz said it last time,
http://www.investorshub.com/boards/read_msg.asp?message_id=1018287

But In my opinion, if your local Best Buy and Costco doesn't have Centrino on the shelf, then I think you need to live in an area that has better stores....

Best Buy Centrino Here: 10 Models Available (Type Centrino in search box)
http://www.bestbuy.com/

Costco Centrino Here: 7 Models Available
http://www.costco.com/frameset.asp?trg=search.asp&srch=centrino

Semi

Dan3, from the article:

Many of the users have gone to the X1000 forum and complained about problems with the keyboard firmware and apparent bugs inn the touchpad.


What does that have to do with the Centrino package, or with a recall of it? I rather looks like bad design work by HP.
Keith


Obviously, it has nothing to do with Centrino, and no Centrino's have been "recalled" as Dan claims. Are you only now beginning to realize that in Dan's mind, and most of his postings, that ANYTHING that goes wrong with a product that has intel inside, is intel's fault? Welcome to consciousness.



Semi

BTW, thanks for pointing an obvious posting inaccuracy. I'm sure jhalada will be joining your comments pointing out Dan's inaccuracy momentarily........ Not.

Hey Beemer, Check This Out....

http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=19221047
To:Tenchusatsu who wrote (175180)
From: steve harris Monday, Aug 18, 2003 5:50 PM
View Replies (1) / Respond to of 175188
Tenchusatsu,
But AMD has more potential!
lol

Can't you get the ihub boys back? I miss the good old days.
Steve


Boy, can you beat that for irony. After what happened there, and from one of the main people who was probably instrumental in getting Paul Engel Banned, instigating the subsequent exodus, which resulted in the creation of the SI Moderated AMD Cheerleader Thread.......

Steve Harris wishing.......

How come nobody told me Hell Froze Over?

Semi

thatsa causioso bon granda lafftisimo en ma belliosa.:)

Me alsomio grande laffo mi assimo offalo....



Semi

A NEW OPERA!!!!

(Copied From Another Place)

NYC - This fall, the Metropolitan Opera is scheduled to produce an opera commemorating Bill Clinton's experiences from 8 years in the White House. Composed by Giuliani Veritas (in Italian), it was commissioned by Jesse Helms in conjunction with the National Endowment for the Arts.

"LA BUBBA VITA"
by Giuliani Veritas

Act I. The Situation: Bill Clinton has been elected President of the United States by an overwhelming margin. The Republicans are devastated, angry and are trying to find their way back to power. As the curtain rises on the opera, the House Republicans are meeting with Ken Starr with the object of trying to find a way to remove Bill Clinton from the Presidency.

The opening chorale, "We Must Finda Way" (Creato grandissimo floozi scandala) is sung as a sextet. In an impressive recitative, Tom DeLay sings "Where Will We Find a Helper" (Dredgi uppulia una Granda Bimbo).

The House Republicans exit. Paula Jones enters stage right with a mirror, singing her plaintive "Why Can't I Find a Man?" (Mia schnozola es humongo.) Tom DeLay and Newt Gingrich enter from the other wing. They spot Paula and sing the duet "Why Not Her?" (La floozi perfecta). They meet and take Paula to a small cafe where they hatch their plot in hushed tones.

Paula tells them of her meeting in a hotel with Clinton years earlier and how her fortunes have collapsed since then. DeLay and Gingrich offer to help. They sing the aria "Your Luck has Changed" (Nozjobbo e'rewardo).

Act II. The House Republicans reconvene with the news of Paula's revelations. They sing in jubilation "We Must Tell the World" (Fono tabloido). The rear curtain raises toreveal the Chorus of Media who sing the chorale "Tell Us More, But Only the Truth" (Sexio scan! dala hypo sweepi). Gingrich enters with Pat Robertson. They sing the duet "He Must Go" (Hypocriti pious crappola). Robertson offers to make time on his television program to expose the charges. At the House Republicans' suggestion, Paula initiates a lawsuit. The Paula Jones scandal becomes the topic of conversation throughout the country.

The Chorus of Lawyers enters from the right to sing the jubilant grand chorale. "We Must do Our Duty" (Multi, multi grande moola).

Ken Starr meets with the House Republicans to plan the next steps. They sing the aria "We Will Save the Country" (Sleezi connivo, la media succisttuppo).

Starr promises to convene a grand jury which will send charges to
the Congress. He sings, "The Truth Will be Known" (Whitewater non
starto, il probo la floozy epidemio). The Chorus of Lawyers sings a reprise of "We Must Do Our Duty" (Multi,multi grande moola!) as the act ends.

Act III. Linda Tripp enters the stage arm in arm with Ken Starr.
She is wearing a headset and singing "Monica is My Dearest Friend" (Io sono la wiccida witchi occidenta).She tells Starr about the secret tapes that she has made of conversations with Monica Lewinsky. Starr takes them from her and sings "We've GotHim Now" (Presidente droppo pantaloni).

Starr hurries off to the Grand Jury to call Monica as a witness. In Scene 2, Monica enters the grand jury room where the Chorus of Lawyers asks her questions. They sing the recitative "How Did It Happen?" (Panti thongo, la flashi). Monica sings the long passionate aria "We Were Meant for Each Other" (Nonsmoko El Pruducto, Phallisymboglio).

In Scene 3, Hillary and Bill are sitting in the Lincoln Bedroom talking about the revelations about Monica. Hillary sings "I Will Stand By You" (Tu jercho estupido, io removo tu equippamento). Bill replies with "She Was the Only One" (Non conto Gennifer, Paula, piu multibimbo forgetta). They embrace!

Act IV. Sam Donaldson is interviewing Henry Hyde in the Capitol
Building. The Chorus of Lawyers hums in the background. Hyde sings the aria "We Believe in Something" (Impeaccho hippi bastardo). Donaldson sings a recitative in answer, "We Only Want the Truth" (Toupee eslippo).

The great trial begins in the Senate. Trent Lott reacts to public opinion polls showing that the President has a 76% approval rating with the public with the poignant aria "What is Right is Not Popular" (Partia repubblico commitini suicido). The Chorus of Lawyers sings the chorale "Principles Come First" (Mi adultero nonconto). With great flourish, Henry Hyde, Bill McCullom and Tom DeLay stand before the Senate to present their case. They sing the somber trio "How Can You Not Convict?" (Evidenso multiflimsioso).

Finally in a moving chorale, the Chorus of Lawyers sings "For the
Good of the Nation, We Must Acquit" (Senatorios non stupido.) After the vote is announced, Henry Hyde, Tom DeLay, Trent Lott and! Bill McCollum leave the Senate Chamber singing the grand quartet "We Still Know the Truth" (Wasto multi millioni) as the act ends.

Epilogue. President Clinton sings the contrite aria "I am Very
Sorry" (Revengo futuro furioso) as the Chorus of Media circles him, shouting their questions. They sing "Who will now Believe us?" (Publicca desgustanta es in media).Monica Lewinsky strolls across the stage on the arm of her new literary agent, Ken Starr. They sing a stirring duet, "It is Still NotOver" (Publishi grande bucchi, conto multi, multi dollare millioni) as the curtain falls.

FINITO

http://www.investorshub.com/boards/read_msg.asp?message_id=1333122

Speculation On Broadcom Takover By INTC

http://finance.yahoo.com/mp#intc
10:04AM BRCM follow-up : Source of acquisition rumor appears to be comment by Piper Jaffray analyst Ashok Kumar. The analyst comments that a hypothetical acquisition of Broadcom (BRCM +5.7%) by Intel (INTC) would represent minimal organizational integration risk With ServerWorks likely to get squeezed out in 2004, Kumar believes BRCM faces a rough road ahead as a stand-alone co. With a mkt-cap of $6.3 bln, BRCM could be acquired for less than what Intel has spent on its previous communications acquisitions. Piper Jaffray notes that this is a hypothetical analysis and that it has no insight into Intel's acquisition plans.

I find it hard to believe that there would be a marriage of these two, considering their recent court cases, but who knows???

Semi

MiTAC set to launch first Intel-Microsoft phone
Tuesday August 12, 4:45 pm ET
By Lucas van Grinsven, European Technology Correspondent

(Adds comments from industry source and Intel)
AMSTERDAM, Aug 12 (Reuters) - Personal computer heavyweights Microsoft (NasdaqNM:MSFT - News) and Intel (NasdaqNM:INTC - News) will have their first joint mobile phone hit the market this autumn, as shown to consumers by Taiwan's MiTAC (Taiwan:3005.TW - News) on Tuesday.
It is also set to be the first mobile phone to use a core chipset by Intel, the world's largest semiconductor maker. Samsung and Motorola will use Intel chips in certain mobile phone models due towards the end of the year.


This is a big win for intel, in trying to move outside of PC/Server chips. New markets means new revenue. Good for them.

Semi

"Advertising allowances are based on a schedule that relates to units purchased, ie volume."

Huh? Who told you that?


I don't know about Dell, but that's the way it works for resellers....

http://www.intel.com/ca/pressroom/2002/0917b.htm
Last year, Intel announced that Intel dealers could use the Intel Inside brand in print advertising and on company Web sites, in addition to broadcast advertising outlets. Dealers are allocated co-op marketing funds based on processor purchases. To qualify for the program, dealers must successfully complete special training and market products containing Intel processors

SemiconEng, Thanks for not escelating the fud, I over reacted.


---Same here, No need for escalation, but I was tempted......


Planerization, (as I understand it) would seem to be subject to the scale effect. I should have stopped there, but here's another angle. Area for 300mm is ~2.25 * 200mm wafer. Assume that there is an optimal size for the current tooling. What might that be? Just on the face of it 200:300 is a large step and in a conservative model presumes optimum is higher.


---Could be, I'm not a Planer Person, but I know surface uniformity is of the utmost concern for Planer people. My guess would be that for larger 300mm tools, the polish pad would also be larger, so it might be a wash, but I can't say either of those, for a fact. I've seen planer tools in operation, and the polishing looks mostly to me, like an orbital car waxer, the pad rotates, but the whole pad assembly also revolves, most likely for the same "Swirl Prevention" as the waxer. Constant Pad pressure is maintained across the wafer surface, so the pad doesn't need to cover the whole wafer at once, and I would "guess" that uniform pad pressure, age of the pad, and pad cleanliness, would be more of a factor than pad size. ALL WAG of course.


Double the size of the wafer, compound that with double device dencity. As YB pointed out large caches may soak up that very expensive realestate.


--- Sure, big cache takes up lots of space, but 300mm also has lots of extra space. Most Modern Cache has redundancy, so if defects spoil a transistor or 2, just fuse it, and use the recundancy. I believe intel does have this chache redundency, and they certainly need to soak up all the "underutilized Fab capacity" (Cough) that Dan keeps talking about, right?


I think it was noted in Bill and Ted's Excelent Adventure, "If such a thing as infinite smallness exists then infinite largeness also exist", can't have one without the other.


---True, for particles, it's good thing those Wet Clean guys can get rid of all that Largeness on the wafer surface. Now infinite small..... that's a pain for sure.

Semi

SemiconEng, My sarcasm was in response to yours.
As far as larger wafers, the larger size is obviously causing some cost. What parameter drives this added cost. What might be the trade offs.
I would suggest that in a perfect world Intel would get wafers that matched their specfications. My guess is that when Intel processes a wafer defects are not exclusively random. I can further speculate that the ratio of yeild to wafer size is not liniar.


Touché then on the sarcasm

The additional manufacturing costs for 300mm are almost all related to the purchase of the equipment, and related to the additional time to shoot wafers in litho, and possibly additional time in planerization/polish. I don't think that Film Deposition and Etch would be as affected by wafer size, since the entire wafer is dep'd or etched at the same time. If a manufacturer paid a little more for the 200mm equipment, they could have purchased a 300mm "capable" tool, lowering the need to purchase new equipment.

As far as defects go, Elmer is more of an expert than I am, but from what I've seen, defect density is actually lower on 300mm wafers than 200mm wafers, so the reported increased yields could be a by product of that lower DD.

Semi

Maybe if your an Intel investor, all those fabs with big wafers pumping out hot P4s, making the above statement maybe more for self reasurance than you maybe realize, maybe.
Someone might speculate: "The contractor supplying the wafers are loosing $ trying to manufacture to technically difficult specifications and further escelating prices could develope..."
That might reflect badly on our beloved Intel. Glad I don't make such speculations.


Your sarcasm, and obvious lack of knowledge of semiconductor manufacturing reveals plenty. Trust me, the silicon suppliers are well able to make 300mm wafers, and my "Experience" in semiconductor manufacturing, shows that silicon costs go down with time, not up, as you "speculate". The cost difference betwen a $50.00 wafer, and a $500.00 wafer can be made up by getting 1 additional 3.2GHz P4 die at end of line. That, in conjunction with the lower 30% cost advantage makes it worth it........ to my "beloved intel" at least. How silly of them to go for higher yields, higher output, and lower costs, so they can drive down prices, while maintaining profit margins. Not "Smart" like our "Negative Profits", "Beloved AMD", eh?

Semi

Prices for 300-mm wafers remain too high, says analyst
http://www.siliconstrategies.com/story/OEG20030620S0020

A bit of a late reply, was away for some time, just thought this article would be worth discussing. While 300mm is definitely the way to go in the future, AMD being limited to 200 mm might not be much of a disadvantage at this point in time.
Keith


A 300mm wafer cost 10 times as much? So what. A bare silicon 200mm wafer goes for about $50.00. That makes a bare silicon 300mm wafer $500.00. With the yield and output number increases described below, it still seems to make sense for a high volume manufacturer. If you're a high volume manufacturer like intel, the cost benifits look a bit different. I guess that depends on how many die you're making, and whether you get a volume discount on buying more wafers....

Maybe, if you're taking into account the relatively low volume of chips manufactured by AMD, maybe it doesn't make sense. Of course, if you don't have any 300mm equipment, or you've been consistantly losing money, or you had to go begging somebody else to make your chips, then yeah, I agree, you may not be able to do it. But to say "it's not much of a disadvantage" to not to be able to increase yields, and increase throughput..... While Reducing Facility Costs........ I don't buy it.


http://www.intel.com/research/silicon/wafers.htm
Cost Benefits of 300 mm Wafers
The primary cost benefit is the ratio of die yeild per 300 mm wafer to die yield per 200 mm wafer. The yield improvement ranges from 2.25X for small die to 2.5X for larger die. The next cost benefit is quantity of output. The output of a 300 mm factory is 1.6-2.2X that of a 200 mm factory producing the same product. The output ratio also reduces the impact of the facility cost on product. A 300 mm factory will cost close to the same as a 200 mm factory but the product output is higher. The die and output scalars are the key drivers that provide an opportunity to achieve a 30% cost reduction for product made with 300 mm wafers.

Semi

Mercury research claims that AMD lost marketshare in Q2 comparing to Q1. How do they count? We know that inventories fell in Q2, so if they count shipments from AMD to channels they must be down. If they count actual sales - no good.
http://zdnet.com.com/2100-1103_2-5059231.html


I believe Mercury Research counts sales the same way that Intel and AMD count sales, when the Distributor sells it, not when the manufacturer ships it.

Semi

Q3/Q4 rally this year?

In 5 of the past 7 years, including 2001 and 2002, INTC rallied during the second half of the year. Given the run-up we've just experienced in the past 4 months, what's the likelihood that we will see a repeat performance in 2003? Price targets for the end of the year?


So, I don't know about Q3, but I've been following this "secret strategy" for the last few years, in which during Q4, I dump my ENTIRE portfolio into intel at the beginning of Q4, and divest it back out at the beginning of the following Q1. I keep an excel spreadsheet of the results, and Here's how it's turned out. Start, is the stock price on Oct1 of each year, and finish is the stock price on Jan1, 3 months later:

Year Start$ Finish$ %Yearly Avg %Running Avg
1998 83.87 118.56 41.4 41.4
1999 74.94 82.31 9.8 25.6
2000 39.94 30.06 -24.7 8.8
2001 21.96 31.45 43.2 17.4
2002 14.05 16.68 18.7 17.7

OK, so the table didn't come out so good, but you can see, that granted, watching your investment increase 41.4% like it did in 1998 is a reason to do a "Happy dance", but on the other hand, watching your investment drop 24.7% like it did in 2000 is enough to give a person a heart attack. Still, since the avg return since 1998 is +17.7 percent, and when you take into account that that return is ONLY during the last 3 months of the year, I'm satisfied with that strategy...... so far, and I plan to do it again this year.

Semi

according to a report from investment banking firm American Technology Research Corp

I would venture to guess that an "investment banking firm", that nobody's heard of, doesn't know squat about what's really going on inside intel.

Semi

OT drjohn, apologies to the board ahead of time


http://www.investorshub.com/boards/complex_terms.asp
A Personal Attack is defined as:

Posting harassing or otherwise objectional content on another poster;
Calling another poster names or being vulgar;
Not staying on topic with the current investment discussion, but instead focusing on an individual poster;

AMD Preparing an Opteron Paper Launch??

http://www.overclockers.com/articles791/
AMD has projected huge rampups before, just to have them vanish. Even if you're not skeptical about that, the production numbers for the 1Q seem to indicate continued big problems in making the things. Given the not-too-bad shipment numbers for the simpler socket 754 chip, this would seem to indicate some sort of yield problem with the more complex chip. They can make some, but not a lot.

It seems to me that somebody's getting excuses ready for a little-more-than-paper launch and early trajectory of the more complex Athlon64.

Semi

Wireless or ethernet plugin at the meetings. I assume wireless.

I wish wireless. The Ethernet wiring was installed almost 3 years ago, before wireless was widespread.

Semi

AMD Preparing an Opteron Paper Launch??

http://www.overclockers.com/articles791/
AMD has projected huge rampups before, just to have them vanish. Even if you're not skeptical about that, the production numbers for the 1Q seem to indicate continued big problems in making the things. Given the not-too-bad shipment numbers for the simpler socket 754 chip, this would seem to indicate some sort of yield problem with the more complex chip. They can make some, but not a lot.

It seems to me that somebody's getting excuses ready for a little-more-than-paper launch and early trajectory of the more complex Athlon64.

Semi

Semi - I think it ended up as intending to be an eyesore. After what the Texas State Legislature did, I think intel wanted to remind them of it for awhile.

What did they do?


So, here in the Valley of the Sun, there are 4 semiconductor manufacturers, so you frequently hear gossip from the other companies. The way I heard it......

Intel was deciding where to put this "Development Building", and secured financial concessions from the Texas legislature (as companies always do), to build it in Austin. At the time, they were gung ho to get High Tech in that area. The stupid thing I heard intel did, was, that they didn't get it in writing. They bought the property, cleared the land, and started erecting the structure.

Then, the Texas legislature developed a case of "amnesia" about the concessions. Noooooo says Texas, we never promised any concessions, what are you talking about? The story goes that Texas thought that since intel had started construction, they would NEVER put it on hold. Circumstantial evidence seems they did.

The reason it got to be a story around here, was that when thy were putting up Fab22, there were several stories going around about the steel intel used to construct the F22 Office Building's superstructure. All the steel arrived on the site at the same time, and according to a TV News story I saw, intel shipped all the steel by rail....... From Texas.

This was a few months after they stopped construction on the Texas building. The funny thing was, I saw the structure when they were building it, and the steel was oxidized, you know, that red rust color? Like it had been sitting around for awhile?

Of course, you know how stories go, and it could be just a bunch of B.S., but if you look at the F22 Office building, and the F12 Office building across the parking lot, it's clear that the designs are different. On the other site in Chandler, when they built an additional office building they made it look like what was already there, not on the Ocotillo site. And guess who I heard occupies the top floor of the Ocotillo building.....

The people that were originally scheduled to be assigned to Texas.....

Coincidence? Maybe. Gossip? Probably. But it all seems to fit. Interesting story though, eh?

Semi

That's a solution, but you are still stuck with sub-par components that go into a typical laptop in order to achieve space and power savings.

Maybe, but I'm hooked up to a high speed network, and since most of the heavy number crunching activities are done on the network server, and power requirements aren't an issue when I'm docked, because I'm plugged in, so I don't need a power saving chip. I haven't found the set up to be limiting for me. Actually, I like it, because all the conference rooms are networked, so I undock, go to my meeting, link back in, and away I go. We have a connection to the overhead projector for presentations, so I haven't taken a pen/paper to meetings or made overhead foils in years. Depends on what your doing I guess.

Semi

Intel sells 'eyesore' structure in Texas

About time intel got rid of it. The thing is, I think it ended up as intending to be an eyesore. After what the Texas State Legislature did, I think intel wanted to remind them of it for awhile.

Semi