Sgolds -

SOI did help prevent leakage into the substrate but there were unexpected problems with horizontal leakage. Apparently, IBM had some solutions to this problem, as I understand it.

The same IBM fab operation that lost money because of yield problems? They may have addressed those problems but it remains to be seen if they solved them.

Sgolds -

180nm copper came up really smoothly

I don't remember that but I do recall that eventually it appeared to work fairly well. Curiously though, AMD never reached their theoretical capacity, even with high demand.

Sgolds -

So we have the vapour sample EE being beaten by the shipping FX, and Intel boosters falling back on comparing against a processor from a different market segment. Desperate times call for desperate measures!

Different benchmarks draw different conclusions. You pick yours and others pick theirs.

I find it amusing that you praise a company that has posted losses for 8 straight quarters and faces more for as far as we can see, meanwhile you characterize a wildly profitable company as suffering "desperate times". Clearly your measure of success is not the same as most others. If success is posting a slightly higher benchmark for pimple faced geeks blasting mutant aliens then enjoy it! Investors will measure their success in a different way.

Sgolds -

I realize that these benchmarks aren't overly relevent. Perhaps you missed it when I said "let's not dwell on topend benchmarks when the volume is elsewhere". But they are an industry standard and despite the sour grapes comments by a few, the Intel compilers are the best in the business.

Any predictions who will win next week's financial benchmarks? I've got a bunch of INTC calls out at $30 and AMD Puts at $7.5 & $10 that are about to expire. I'm waiting for the traditional drop in AMD's share price after earnings to write some more Puts. Some $7.5 Calls are coming up too and it looks like I'm going to have to let those shares go.

Petz -

The "75% later this quarter" seems absurd to me.

Of course it does. You still think it's not pulling it's weight. Move on to your next misrepresentation.

Petz -

P4EE missed the Q3 SPEC results, so its obvious that the results published on Intel's website were not submitted when the product was released, but more than two weeks after its paper launch.

The disclosure says that the results were published in October, 2003.

I understand that the blowaway performance of the P4EE is a crushing blow but please don't misrepresent the facts.

The site clearly says the test results are from September. More than one person has explained to you that the results aren't published immediately but must be subject to review. Get over it and move on to your next misrepresentation.

http://www.spec.org/osg/cpu2000/results/res2003q4/cpu2000-20030922-02523.html

Jerry R

Re: 3.2GHz P4EE

The commanding lead P4EE holds(15%) over anything AMD can muster is about the largest lead I can remember ever seeing. But let us not dwell on benchmarks of topend products when the volume is elsewhere. The only real benchmark that matters will be posted next week by both companies and it will be an even bigger lead by Intel.

P4 EE SPEC scores posted on SPEC website.


http://www.spec.org/osg/cpu2000/results/res2003q4/

P4 EE SPEC scores posted on SPEC website.


http://www.spec.org/osg/cpu2000/results/res2003q4/

Sgolds -

They had better have the bugs worked out before high volume. You can continue to improve bin splits and yields with high volume, but you have to have a fundamentally working process with acceptable yields and bin splits before going high volume. That is the part that AMD missed when they first went to 130nm - and it cost them a lot.

That's exactly the point. They don't know if they even have a good stable process without high volume. Without an HVM process at 130nm they are not positioned to move to 90nm. The low volumes show that either they don't have SOI@130nm running smoothly or else the demand is very low, or both. You can't produce a few 10s of thousands of units and say the process is healthy.

Sgolds -

Right now they need just enough 130nm SOI production to meet demand in a Windows-less world.

Wrong answer. They need high volume to work the bugs out.

SGI doing better than expected. Probably good news for Itanium.

SGI Sees Q1 Revenue Between $215 to $220 Million

Boston, Oct 06, 2003 (MidnightTrader via COMTEX) --

SGI (SGI) tonight said it expects Q1 revenue between $215 million and $220 million, above prior guidance of $200 million to $210 million, and ahead of the lone analyst estimate of First Call of $207 million.

The company expects an operating loss of between $40 and $45 million, including a previously announced restructuring charge of approximately $20 million. Excluding the charges, the non-GAAP operating loss is expected to be between $20 and $25 million.

wbmw -

aHT vrs PCI-Express.

Both are highspeed differential links although PCI-Express is much faster. AMD puts 16 of them together (8 in each direction) to make a channel. What's to stop Intel from doing the same?

wbmw -

Well, that's a pretty unique view.

I don't think so. You have 2 cpus with their own memory and connected by a highspeed link. That describes both systems.

wbmw -

Sorry, but I disagree.

I meant for Opteron. IBM's offerings are essentially clusters of 2-way systems.

Sgolds -

My view of the architecture is that all Opteron MP systems are clusters, even if they are on the same board. aHT is effectively the highspeed link and I don't see any difference between this and a cluster.

Sgolds -

Multiprocessing. Scaling 2P, 4P and 8P will be even more impressive.

I don't know if you've noticed or not but clusters or 2-ways seems to be the way things are going...

chr p

Intel must profit much more money to deserve market cap of 192.37B :)

So why does AMD have any value at all?

BUGGI -

I could imagine what they will say:

What they always say:

Intel "We made a ton of money".

AMD "We lost a ton of money".

ChipGuy -

Well that will be great for two Opterons talking to each
other in an MP system but is there any suggestion that
AMD's chipset partners will be able to keep up?

I wonder if they can run asynchronously, meaning the inter-processor ports at 1GHz and the I/O at 800MHz? Probably not.

SZ -

Synopsys Announces Availability of Galaxy Design and Discovery Verification Platform Tools for Intel Itanium 2-Based Systems

A little birdie told me that one of their largest customers is moving to IPF workstations for their design platform...

wbmw -

are you talking about selloffs besides the chip unit?

It was just a guess on my part. I don't want to speculate too much because I don't know enough about their business.

wbmw -

Any speculation on the possible consequences?

Only speculation of course but it looks to me like they don't want the losses to drag down the rest of the company. Could a selloff be in the future?

The first step to unloading?


Motorola to spin off chip unit


http://rss.com.com/2100-1006_3-5086763.html?part=rss&tag=feed&subj=news

sgolds -

The very definition you offered implies unethical behavior. Money offered to a person in a position of trust to influence their judgement or conduct.

The second definition is so vague as to allow just about anything that involves the exchange of money to be called a bribe. I don't buy it.

Sgolds -

Not all bribery is illegal or even unethical. Bribery is only illegal if it is used to induce someone to break a law.

Boy this is a great circular argument.

Bribery is by it's primary definition a violation of the law or at least unethical. According to Websters, if it isn't illegal or unethical then it isn't bribery. Can you please provide a dictionary definition of bribery that agrees with your concocted definition?

By it's secondary definition, you could call it bribery if you offer a company official or purchasing agent a "personal incentive" to select an HP system when that decision is otherwise in all respects the wrong decision for the company to make. In HP's case they are offering the incentive to the actual company and the incentive then becomes part of deciding what is in the company's best interests.

If this were AMD making the offer there's no way you would be torturing logic this way.

IP -

I believe "bribing" is a proper word.

Yes, it is a proper word but it isn't the proper word here.

Perhaps you should consult a dictionary.

bribe:

1) anything, esp. money, given or promised to induce a person to do something illegal or wrong.

2) anything given or promised to induce a person to do something against his wishes.

Websters New World Dictionary.



Now unless you are claiming that buying an HP system is illegal or wrong, or at the least against a company's wishes, I think the word is incorrectly used here. However, knowing the vivid imagination of some of the posters here perhaps the Inquirer has convinced them that buying from HP, because of their use of Itanium processors, is an illegal act of wrong doing. It wouldn't surprise me in the least.

Sgolds -

I think "bribing" is a little strong of a word. They look like incentives to me. I'd hate to think I was being bribed every time I buy the coffee on sale at the supermarket.

Nevertheless, I think the only remaining question is who will end up with Sun's IP?

Jack -

I assume your assuming that I think that you are wrong

Wrong again. I assume intelligent people wouldn't do what you did, rely on tabloids. I'm still convinced intelligent people wouldn't do that.

Jack -

You assume a lot of things don't you?

Yes, I assume intelligent people don't rely on sources that have little credibility. I still think I'm right on that assumption.

Sgolds -

My last comment on this.

Until the next tabloid rumor comes along...

Sgolds -

please show me where I said I believed the article.

Here:

"So, your argument is with Intel marketing, not with me."

You are clearly giving credibility to the tabloid rumor and assigning responsibility to Intel Marketing.

Smith proposes tax break for firms' foreign profits

The Bush administration says it's skeptical of the projected reinvestment benefits and fairness of the measure

10/03/03

TOM DETZEL

WASHINGTON -- Sen. Gordon Smith is claiming credit for engineering a potential major tax holiday for companies such as Intel, Nike and Hewlett-Packard that even the Bush administration calls unfair and critics call a corporate giveaway.

The Oregon Republican sponsored provisions in a bill that cleared the Senate Finance Committee on Wednesday to allow a temporary, one-time tax break on some $300 billion in profits that U.S. firms have kept overseas, thus avoiding federal taxes.

<more>

http://www.oregonlive.com/metro/oregonian/index.ssf?/base/news/1065182497314870.xml


So how does this affect Intel's earnings? Are overseas profits excluded from the earnings reports? Just how much earnings does Intel have accumulated overseas?

Sgolds -

So, your argument is with Intel marketing, not with me.

A tabloid rumor from an unnamed source. Once again a fine source of reliable data for those who want so desperately to believe.

Here's another excellent source of factual information:

http://www.nationalenquirer.com/

Jack -

So you believe everything you read on tabloids. I assume you believe the abduction stories as well.

John -

Our capitalism is build on accessibility to the consumer. If we start down the road to limit that, it could lead down a road to less prosperity.

So why then shouldn't they be allowed to come into your home and present their wares without invitation? Perhaps they can also follow you into the bathroom? We wouldn't want to limit accessibility to the consumer now would we? In my opinion the home is a sanctuary and not available to uninvited intrusion.

Sgolds -

I think that the obvious is now shown to be true - the fully burdened CPU cost for Intel is way higher than AMD. Thus, Intel could not lower their price to be below AMD without violating anti-trust law.

I guess like some others you're convinced yourself that your argument is bullet proof but I'm not convinced. Remember you've got to convince a jury that you've violated anti-trust law and you guys haven't even begun to address all the differing variables that could be brought up to show that the costs you've included should not all be burdened onto CPUs. What proportion of those Marketing costs should be distributed across the other product lines? What portion of the R&D goes to Flash, Wireless, Comms, networking, automotive, consumer electronics, embedded controllers, military, xscale, chipsets etc etc on and on. What about depreciation? You can't dump that all on Processors either. A smokescreen like you couldn't imagine would make proving anything far too complex to ever succeed.

Jerry R

AMD's introduction of the Athlon 64 demonstrated that they could not deliver a knockout to Intel - at best, they match Intel's fastest offerings and is clearly not a situation similar to 1999 and the original Athlon launch.

Jerry I think it's a mistake to simply look at the Marketing benchmarks and try to determine a company's competitive position. Especially when you are making investment decisions. I would look at a company's financial position and track record before deciding how much faith to put in their promises.

drjohn -

Your post proves conclusively that you don't understand that selective accounting is at the very least misleading and often fraudulent.

You are posting this to the wrong person. That statement was by Petz, not me and I agree with you that he doesn't the issue.

Tenchu -

Oh no. Persuasion and deception are two very different things. One is obviously an integral element of marketing; the other should never be for both ethical and legal reasons.

What about AMD's quantahurtz? That relied on deception to be effective. Are you saying it should have been illegal or was it just unethical?

If a letter came into my mailbox claiming to be from my aunt, but instead had advertising for porn, should that be protected under the right of free speech? Seems like a lot of the junk e-mail out there does something equivalent.

Criminally it would be protected speech but civilly could you show damages?