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.
chipguy,
Given the fact that to a large extent AMD's server
products are worthwile today due to its ripping off
ideas from those "white elephants" such claimed
forgetfullness is more likely feigned amnesia.
It could be real. You can do 9 out of 10 things right, and one thing wrong (of using a wrong instruction set) and you end up on the ash heap of history.
Joe
chipguy,
The fact that no major OEMs ship 8 socket K8 boxes...
There are no major OEM 1 socket Opteron boxes either.
... and there are still no benchmark results for no-name 8 banger systems should tell you something.
First of all, it tells me that the no-name OEMs don't have the marketing budgets to make benchmark submissions (confirmed by lack of submissions for 2 and 4 socket systems). Going further than that is speculation.
BTW, it seems that you are subscribing to "Elmerian logic" too.
Joe
chipguy,
AMD partisans crack me up. To them nothing ever happens in
the computing world until AMD does it.
You have to excuse us. It is easy to forget the white elephants, past and present.
Joe
Alan,
I heard the cache coherency traffic goes up as the square of the number of CPU's... The broadcast coherency method AMD uses runs out of gas once they get up near 4 sockets, which is why even though it currently allows 8 sockets you don't see many of those systems out there.
Elmer used to claim this about 4 sockets, before there were 4 socket system, explaining that absence of 4 socket system is evidence that HT "runs out of gas" with 4 sockets. It seems like you guys are reading from the same script...
BTW, there are 8 way Opteron system out there from lower tier vendors. There are no benchmark submissions from those systems as of yet.
When you make these claims about Opteron "running out of gas" , the question you have to ask is: Compared to what? Compared to hypothetical processor with more advanced cache coherency? Or compared to real competitor - meaning Xeons on shared FSB bus, or actually 2 5 drop FSB buses interconnected by slow connection?
A more sophisticated coherency scheme is required at that level.
There may be perceptible performance degradation using the current scheme with 8 sockets, especially DC that may reduce performance scaling. We down't know. But there is a brute force way to scale up to 8 sockets, maintaining obviously acceptable performance scaling of a 4 socket system. It can be done by adding an extra (4th) HT link, and not changing anything else.
Joe
I don't really frequent DIY stores, but looking online, availability of higher end stuff is great.
As far as A64 3200, I was quoting Newegg price, which is usually a reasobable indicator.
Looking at PriceWatch, I see $162 for Winchester, and $168 for Venice, no $155.
Joe
Bobs,
re: supply issues
AMD used to sell A64 3000, that was priced between $140 and $150 for the longest time. It is not available any more. The lowest speed grade is A64 3200 now, sold for $174:
http://www.newegg.com/Product/ProductList.asp?Manufactory=&PropertyCodeValue=0&PropertyCodeV...
BTW, it is identical die, most likely downbinned only 1 bin instead of 2...
Joe
delete duplicate
Durl,
I have forgotten more about finance than you will ever know.
I guess topics on capital assets and their depreciation is something you have forgotten long time ago...
And graph number 10 shows projected debt for AMD, but number nine of Hyjack park's post seems to show a current ratio of less than one. (cash and AR to debt) Whoopsee.
... and let me add that you don't know how to follow the thread of discussion.
It is kind of like a tree. You can draw it with a pencil on a piece of paper.
You make post with 2 dumb statements, namely, #1:
This is my gut feeling. I am sure I could be very wrong but my gut feeling tells me they need cash of 2 billion this next year and they are making about 150 Million.
and #2:
My guess is they pissed away about 2-4 billion on Opteron which can't seem to garner more than about 12% of one segment of the market.
Instead of substantiating them, you attack a post in another branch of the tree, which is not even mine.
BTW, this is not the first time you did that, therefore my conclusion.
Joe
Durl,
This is my gut feeling. I am sure I could be very wrong but my gut feeling tells me they need cash of 2 billion this next year and they are making about 150 Million.
You don't seem to know much about finance...
My guess is they pissed away about 2-4 billion on Opteron which can't seem to garner more than about 12% of one segment of the market.
... or technology.
I guess this is a good place to learn about both, if you set your heart to it.
I could really do a cash flow analysis of AMD, but I really don't have a need to do so.
Yeah, and I could really design a spaceship that would take people to Mars and back, but I really don't have a need to do so...
Joe
Wow! That was very good!
Joe
Keith,
The question I´m thinking of is the one of lost sales. I hope they won´t have AMD to blame for that and actually rely less on them in the future again.
If AMD can't supply any more, than these are lost sales only in theory.
As far as AMD not delivering to HP what was promised, I find that highly unlikely, since they delivered a single CPU to me today (via Newegg) and I am the smallest customer buying one vs. HP, the biggest customer. AMD would surely supply HP, to fulfill its obligations before selling one to me.
That´s one of the "other options".
Another one of the "other options" could still be some "conditions" that HP has to fullfill to some of the illegal Intel monopoly tactics. If it could not be done by customers making selection themselves, than HP may have had to make the selection for them by withholding the choice...
Just the situation per se is not good - customers that want to customize AMD-based systems from HP for now are out of luck, unless they can afford one of the two remaining top-end lines. And this in the Christmas run-up.
At this point, the Christmas runup may be done for the BTO lines. The lead times to build the system and to ship it probably does not leave enough time to sit by 24th. Ready to ship PCs may still be a possibility.
BTW, there was this whole schedule that I noticed on HP site, which said, that to have a delivery by Christmas, the order needs to be placed before certain time...
I think now the Christmas runup, for the late shoppers, is shifting to what's left in retail stores.
Joe
Keith,
I only hope that this is due to HP underestimating demand for those lines, and not one of the other options.
If HP underestimated demand, hopefully they will order more in the future. It is much better to sell out what your have than to be left with unsold inventory.
The same applies for AMD. I hope that AMD will be exiting the quarter without much inventory of finished goods... There are some indications that this is happening. First $1B quarter for CPUs in history...
Joe
Facing high AC bill, University of Florida replaces Xeons with dual core Opterons:
http://www.informationweek.com/story/showArticle.jhtml?articleID=174918309&tid=5978
Rumor has it that University of Alaska is more interested in Xeons...
Joe
Keith,
If I were to give you a choice to pick. Would you chose that Intel based computers sold out, or AMD processor based computers?
Joe
cd,
I'm reminded of Elmer's dance around not being an employee or contractor for Intel. Much like him, Olga "provides services" to Intel.
LOL. Good one.
Joe
PS: I submitted a comment on a comments page. The posts are not live. I guess they are screened. Maybe by Olga herself. Somehow I think my comment will not make it there anytime soon (or at all).
Joe
Tenchu,
Apparently, according to Groo, Oblivion will be the first game to make Xbox 360 shine.
BTW, I have seen the game advertized on a NYC bus, some 3 months before release. Some thing struck me about that as odd. Maybe Groo may be on to someting about the impending crash of the game console market...
Joe
herb,
Gee, I see you are forming a protective circle around Olga...
Joe
Tenchu,
Jim Cramer is another "INTC whore." I don't know why this thread is so hung up on Olga when Cramer has a much wider audience and seems to be even more "blatant" in his pro-INTC "bias."
You don't seem to recognize the distinction between investor, investment analysts, people who trade, pump, try to make gains on stock for themselves and their clients vs. a technology reporter for a business publication.
Here is how she describes herself:
Olga Kharif is a reporter with BusinessWeek Online, where she reports on semiconductors, wireless, the Web, and all sorts of weird tech that comes her way: She's been known to write about coffin furniture and flying cars. Kharif has been with BusinessWeek since 2000. She has a graduate degree in journalism from Medill School of Journalism at Northwestern University, and an undergraduate degree in accounting.
http://www.businessweek.com/the_thread/techbeat/techteam.html
Joe
Tenchu,
Anyone want to tell me if Fleck or Tad LaFountain is an AMD whore or not? No evidence, but like you said, "Lack of public evidence does not preclude the possibility that there is personal gain."
Evidence of personal gain, and the same for their client would only enhance their reputation as stock pickers.
on the other hand, underhanded pumping of a stock, smearing of competition, behind the veil of objective reporting is, IMO, unethical.
Joe
herb,
That would be an interesting case:
"I am not an Intel whore" declares Business Week reporter. <g>
Joe
Herb,
re: "While we don't have evidence of personal gain for Olga"
Then by your own definition She is not a whore.
Lack of public evidence does not preclude the possibility that there is personal gain. Given her writing and how she is willing to sacrifice her reputation for benefit of Intel, there is a probability of personal gain.
Or, maybe she is not prostituting herself. Maybe it is genuine love.
Joe
wbmw,
E. You resort to profanity and attacking the author rather than the content.
Is it your opinion than that the term "whore" should be used only to describe those selling sexual favors, as opposed to those selling one's integrity?
Here is what dictionary.com says:
whore ( P ) Pronunciation Key (hôr, hr)
n.
A prostitute.
A person considered sexually promiscuous.
A person considered as having compromised principles for personal gain.
intr.v. whored, whor·ing, whores
To associate or have sexual relations with prostitutes or a prostitute.
To accept payment in exchange for sexual relations.
To compromise one's principles for personal gain.
http://dictionary.reference.com/search?q=whore
While we don't have evidence of personal gain for Olga, we have a plenty of evidence of compromise of the principles of journalistic principles.
Joe
Tenchu,
My French is really bad, but I'm pretty sure the name of that airport wasn't pronounced "Oh really?"
It is (or was) Orly.
Joe
wbmw,
Maybe they have been making money, but have chosen not to make specific mention of it.
That must be it.
Joe
I_banker,
tecate seems to comes here to seek flame wars more than for the technical or investment info...
Joe
Intel has not been making money on flash either. If they did, they are keeping it a secret.
Maybe we will hear from Intel when they start to make money on flash.
Joe
wbmw,
Any reasonable person would instantly take you for a fanatic and ignore any reasonable comments you make.
You can go ahead and do that.
Joe
herb,
Good attempt at changing the subject. I just found it funny what article you called open minded...
Of course, there is something that an investigative journalist (rather than Olga, the Intel whore) may want to look into. How is it that in a booming market, where capacity is constrained, deliveries are being stretched, nobody is making any money?
Joe
Here's another article regards Spansion spin on the upcoming IPO.Seems to be a pretty open minded article.
Spansion's Expanding Losses
Hardly open minded with a title like that. Spansion losses have been shrinking and are likely to shrink again next quarter. The sequence for 2005 is 110, 90, 50, ??. Are these expanding losses? Olga must be the only person in the investment world to use cumulative quarterly results.
But as Intel whore, she is expected to perform acts that others find less than sensible.
Joe
chipguy,
That is because I could write a frigging 200 page book on
this topic because the implications of relative PFET and
NFET transconductance in MOS processes has significant
and far reaching implications going off into many different
areas of digital and analog MOS circuit design.
Since this is an investment board, I think most here are looking for level of detail that is, let's say 10,000 ft view. Not 30,000, and not view of walking on the ground.
Joe
chipguy,
A real "chipdesigner" wouldn't be so confused.
He has contributed more on the subject than you have so far. Nobody knows all the answers, but any effort to get the answers should be applauded.
Joe
Chris,
LOL, that's a good chart. I think the forecasters have given up on Itanium, as to not continue to make fools of themselves - and still not anger Intel...
Joe
avatar,
I would assume that since AMD is retaining 2/3 of the ownership, so would Fujitsu. So Fujitsu will likely have 26.67%.
Joe
AAMDD,
You really have to wonder about the judgement of these analysts. AMD has been demand limited (in an over-capacity) situation for every quarter from Q4 2000 to Q4 2005. So what is new about overcapacity in 2006?
BTW, IMO Intel has been in an overcapacity situation forever as well.
Joe
UpNDown,
I imagine Dell and all the others OEM (some of whom are bigger than Apple) would not be happy about that.
But it is possible that Intel will pull in the schedule for everyone.
Joe
muzo,
I wonder if there is some problem with my proxy server caching things incorrectly. But this is really screwed up. I could see it getting confused with a newer version of a file, but reporting that a file is missing from an FTP server is something I have never encountered before...
Joe
Keith,
Weird. I go to ftp directly, I try the link and I get nothing. Can you e-mail it to me? Do you still have my e-mail address?
Joe
Keith,
The file does not seem to be there.
Joe
Buggi,
Now that you mentioned this, it may be the OS that is holding things up (not to mention the apps).
What I don't understand is why wouldn't Apple release only a 64 bit OS, and forget about the legacy Yonah CPU?
There are millions of 32 bit (and 16 bit) Windows apps which makes it possible for users to stay with 32 bits without really missing out on much. But Apple is starting with clean slate on x86, so why pollute it (and its OS) with 32 bit Yonah?
Joe
Keith,
Apple said earlier that the company would deliver new Macintosh computers using Intel microprocessors by June...
...A November 30 Reuters report cited Citigroup analyst Richard Gardner as saying that Apple may launch its first Intel-based PowerBook as early as January.
From the dates, it sounds like Yonah. A little weird, if you asked me, to come out with a Yonah based computer almost 6 months after release of the processor. Time to market is important, which makes me wonder about the delay.
If Apple is going to wait until June, why not wait until Merom / Conroe ships a few months later? If Apple is starting fresh in this x86 market, why complicate things with introduction x86-32 legacy, and not go directly to x86-64? For once, Apple could have an instruction set that would not change on semi-annual basis...
The only sound explanation of this is if Intel ships Merom / Conroe early, in June.
Joe