Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Petz -
Sorry, wbmw claimed Cost per CPU was lower at Intel. My post proves conclusively otherwise. AMD's cost per CPU is less than Intel's. Period.
Sure it is Petz, and you were abducted last night too weren't you...
8-/
Can you do the math for us so we can see what you mean?
No, but I'm not the one making illinformed estimates on Intel's volumes, ASPs, cost structure, cash flow and balance sheet.
If you remember the question was if Intel could kill off AMD if they chose to. Then it evolved into how long they could last if they closed unprofitable divisions. Then it drifted into writing off paper losses (which have no effect on cash flow) and ignored the effect of depreciation on cash flow (none) as well. With $11 Billion in the bank I would expect Intel could post $1 Billion losses for a lot longer than 11 quarters. That would be many times long enough to kill off AMD. It would also be very stupid. Vendettas are part of AMD's culture, not Intel's.
Ten -
Do you think e-mail marketeers have a right to flood my mailbox with ads for curing erectile dysfunction (or whatever other product they're trying to sell)? Do they have the right to try and fool anti-spam software, thereby bordering on fraud in a sense?
Deception is an integral element of one aspect of marketing. It is intended to pursuade you to buy something you would otherwise not buy. Junk mail appears in your mailbox just like spam. The senders do everything possible to get you to open it, just like spam. They make it look like something official or maybe make it look like there's a check inside. All deception. I don't see any difference between the two.
Petz -
Intel numbers and TECchannel P4EE numbers -- neither are on spec.org, and the SPECint numbers differ by >100, so the numbers don't count
Maybe not to you but that still leaves the rest of the world.
When Intel submits SPEC scores they submit the hardware configuration, software revs, drivers, compiler versions and switches, in short, everything. SPEC verifys the submission and then posts it. If you think they do that instantly then you don't understand the process. The fact that Intel posts them means they've submitted then. The rest is up to SPEC.
Petz -
Now, isn't it obvious, that if company A can make money selling a product at $110 and company B cannot, company A, in this case AMD, has the lower cost of production?
This has to win the prize for the most bizarre reasoning I've seen.
Sgolds -
I think you are confusing manufacturing cost with the total cost to run a company.
Maybe he is but remember the discussion was about whether or not Intel could kill off AMD if they wanted to. That gets down to cashflow and Intel could easily do it, if they wanted to. The question is would it be in their best interests? AMD keeps Intel from possible anti-trust action and that's worth the annoyance.
Petz -
Sigh.... I ask myself, why bother?
b2l
Can someone point to an AMD presentation on the 90nm AMD node.
I haven't seen a thing.
borusa -
Screaming forward down the road, eyes wide open, focused inteltly on the rear view mirror.
Droidian slip?
borusa -
AMD, OTOH bet the farm on the work of their engineers and scientists. They have taken on debt to give birth to a processor that Intel can only answer (so far) with large cache, thermal challenged, in the future launch. Though, our frineds across the way would disagree. Maybe this is what they call interesting times?
It is interesting to note that you gauge Intel's response only in terms of your own measurement system. To me, Intel makes tons of money and AMD loses tons of money so I have a different perspective on the effectiveness of the "response", if there is even such a thing. Why not let the market decide on the correct measure? They seem to be less impressed with benchmarks and paper releases.
Sgolds -
I think it is too late for Sun, they are too rapidly losing the market to Linux.
I have yet to see anyone make a reasonable argument against this. It just seems inevitable. If we can see it so clearly then why can't Sun? Is this denial on a Corporate level?
Paul -
what's with all the "drive-by" comments tonight?
Hey I'm in a mood... Can't kick the dog or beat the Wife...
8-/
When I do the math it looks to me like INTC made about $30/µP last Q.
Can you do the math for us so we can see what you mean?
SZ -
Please explain this "lower cost per employee" metric.
Obviously AMD underpays their employees. I guess you get what you pay for.
Petz -
Whatever happened to P4 EE SPEC results? Nothing yet submitted
How do you know nothing is submitted?
borusa -
I suspect that AMD's Dresden fab is cutting edge in all aspects except 300mm.
Cutting edge? I guess that explains the low output.
Sgolds -
Yes, AMD has been losing money and Intel has been modestly profitable. Now, at the slim profits that Intel has been making, it should be pretty obvious that if you replaced Intel's ASPs with AMD's ASPs then you no longer have a profitable company. In fact, since their cost per employee is higher, I speculate they would be in worse shape than AMD.
So? They have an entire division that loses money. That could go, and they have large depreciation charges which do not impact cash flow, not to mention cash reserves. They could kill off AMD easily if they wanted to. Obviously they don't see that as being in their best interests.
Sgolds -
I don't agree with this statement once you factor in corporate overhead. Last time I looked, AMD had a lower cost per employee.
We're not talking about producing employees. In case you missed it AMD has lost money in the last 8 quarters straight. A $20 drop in ASPs would cost AMD about $100 million a quarter. That would be death and Intel could keep it up indefinately if all they wanted to do was kill off AMD.
Sgolds -
Where does Sun go from here?........How do the differentiate themselves from every other x86-64 system out there?
What other Opteron systems out there? IBM's is essentially a cluster of 2-way workstations. Opteron on Solaris in larger SMP systems would be a unique offering. Going to standard hardware and software, Linux and Windows, means they're just another vendor and they don't have experience competing on price in a commodity market.
Jack -
Now I understand why it's "so obvious" to you, and only you, it's your speculation.
Looks like that one went over your head. I'll try and explain, the fact that Sun should use Opteron is obvious. The reason why they don't is only speculation. Try and see the difference between the two.
YB -
I think that's right. All those points are correct. There is a bunch of laws that prohibit Intel from completely crushing AMD, but just keeping them low is very expensive.
Clearly Intel can produce cheaper than AMD. I know of no law that would prevent Intel from dropping their prices to a point where they would still be profitable yet AMD could not match them without facing a quick bankruptcy.
From second half of 2000 and almost for a year after that, AMD kept the performance crown, having the Athlon running faster than fastest P4. Intel spent close to 10 $billion in capex to make P4 faster than Athlon
The bulk of that went for new fabs and those would retain their value P4 or no P4, so you are really exaggerating here.
Pay attention that this is 3 times AMD market cap.
And they built about 3 times as many fabs as AMD has.
YB -
Sun is still bigger than Dell in Q2 this year, third after IBM and Q.
The thing is though that Sun is in a death spiral and they aren't going to get out of it without a major change in their business model. SPARC isn't going to keep up with either Intel or AMD. Sun needs to cut their losses and pick one or the other. Going with AMD would offer a more unique solution rather than going with Intel where they'd be one of many.
Jack -
What, exactly, is 'It?. A merger? Sun's adoption of Opteron?
Speculation on my part but it may be that either SUN is not convinced AMD can deliver Opteron in stable quantities or someone else may be offering a more compelling alternative.
JH -
I doubt Amd will be buying Sun anytime soon - Amd's Market cap. is about 4 Billion to Sunw's 11.
Maybe you're right. How about a merger instead? If SUN doesn't select Opteron then that's a telling story. It's so obvious that if they don't do it there must be a very good reason.
Sgolds -
A while ago we were talking about SUN buying out AMD. Maybe AMD will buyout SUN?
blauboad -
That is basically what Walmart does. Sometimes they even price their goods below cost to eliminate competitors.
AMD has been pricing their products below cost for years now. It hasn't eliminated Intel.
C -
Yes, they could do exactly what you suggest, they could kill AMD in short order. That would be the actions of an unethical company on a vendetta instead of an ethical company with the shareholders interests in mind.
C -
Hypothetically, I see no significance in Intel visiting a perspective customer. I don't remember anyone speculating on why AMD might have been seen visiting Dell, or HP or IBM or anyone else. I also place zero credence on tabloid speculation on these matters.
C -
What Intel did doesn't make sense.
How do you know the incident even took place?
Sgolds -
Why am I stating the obvious?
I don't know. Even the most die hard conspiracy theorist couldn't spin this one very far.
SZ -
AMD just stole this technology from Saifun
Well well well.... Isn't this interesting. I wonder just how the Apologists will explain away AMD stealing technology?
Your link didn't work for me but here's another one.
http://aol.thestreet.com/tech/themarker/10035049.html
Here's another that claims AMD lied to Saifun then flat out stole their technology:
http://www.internetnews.com/infra/article.php/1436901
The lawsuit, filed by Saifun In February of 2002, alleged that the company disclosed "revolutionary" memory chip technology to AMD and Fujitsu after the companies agreed to keep the information secret. According to the suit, both companies used the technology to develop their own memory chip products. Saifun had sought an order blocking the companies from infringing its patent.
C -
Yasoo Constantine - welcome aboard.
Errata lists don't play into that but it irks me when some people try to pass AMD as an inferior designer of cpus when they are clearly not
I certainly agree that AMD is a fine designer of chips. Some of them become good products too.
Sgolds -
It seems to me the Calif. SC's decision said essentially that just like junk mail, anyone has the right to contact you at your email address. Whether or not you read/block it is your choice but up to, and including delivery to your box, is protected speech. In regards to the "don't call" list, I suspect the same argument will apply. Your phone is a portal to the outside but it isn't outside. They can call you but you don't have to answer.
wmbz -
I'll have to say that the fact that the Athlon had an errata sheet 1/10 th of the P3 and P4 never sat well with them and they always "suggested" that AMD was not forthcoming with all the errata.
Please don't involve me in discussions I have not joined. Regardless of the length of the errata list, either public or under NDA, the final measure is profits and AMD's list has no entries.
BTW, what other alias do we know you by?
Sgolds -
There is no rational reason why a design team, with enough time to develope a processor, would divide it into two discreet components and then stack them.
Sure there is. 2 die spliced together would yield more than 2X higher than 1 die twice the size. Additionally it suggests the possibility of stacking a third, or a forth, or ....
Spokeshave -
made the claim that functional errors that affect the performance of a processor are always exposed eventually. It is an intuitive statement, and unprovable either way.
So it is a belief and not a statement of fact as originally presented. Opinions are fine as long as you don't try to pass them off as fact as you did.
Sgolds -
The point is simply that both would involve pad to pad contact, that's all. Vias and contacts.
Spokeshave
I eagerly await your response.
I'll bet you do.... The burden is not on me to disprove your unsubstantiated claim. The burden is on you. If you want to qualify it with "in my opinion" then that would be fine but otherwise you made a claim you can't possibly know.
Sgolds -
The word kludge comes to mind! Has theInquirer (or Intel) gone off the deep end?
That's odd... We saw a news article last week about SUN doing a stackable design that sounds very similar yet you never said a word. Strange...
Spokeshave -
Errata that have impact on performance are *ALWAYS* exposed eventually.
Really? How could you possibly know this?