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.
For the previous four years, AMD outpaced Intel in receipts, so its interesting that such a dramatic reversal has taken place.
---More than dramatic. You'd almost think that they had a New Process or something. I would guess, that intel must be doing quite a few things, quite a bit differently, in 90nm over 130nm, no? Doesn't give me the impression of a Dumb Shrink.
***********************
By the way, message for any "Newbies" around, this is one of those little pieces of data I was talking about, the ones that I put together into the bigger picture? Just thought I'd let you know, just in case you couldn't figure it out. No need to thank me for the heads up. Back to our regularly scheduled programming.
***********************
Intel increased their number by 48% over 2002, becoming one of the top 10 organizations receiving US patents. Without seeing the list beyond the top, I suspect that AMD didn't drop significantly (perhaps they increased as well), but it appears that Intel has really stepped up in their filings.
---In any case, I don't see AMD in the top 10. I can't recall who it was on the other threads crowing about the number of AMD patents over intel. I'm guessing we won't be hearing too much crowing about patents this year though.
Semi
More Info On intel TV Chips
I hadn't seen the technology described in detail........
http://www.intel.com/pressroom/archive/releases/20040108corp_a.htm
INTERNATIONAL CONSUMER ELECTRONICS SHOW, LAS VEGAS, Jan. 8, 2004 - Intel President and Chief Operating Officer Paul Otellini today demonstrated Intel silicon technology that is expected to bring about high-definition, large-screen TVs with clearer pictures than current systems for less than $2,000
The new Intel technology, code-named Cayley, is based on a technique called Liquid Crystal on Silicon (LCOS), which is used to create small chips called microdisplays that produce images that are displayed on large-screen, rear projection TVs.
---Better performance for a significantly lower price. Going for that whole price/performance thing I keep hearing about, I guess. Sounds like a winner to me. When people are buying big screeens for ~1k someday, I hope they remember who to thank.
Semi
I realize that Prescott is not out yet, and I reserve the right to change my mind and investment strategy at any time, but for now AMD looks to be the better play for 2004.
HailMary
Sounds reasonable and fair to me. The best of luck to you then.
Semi
http://www.aceshardware.com/forum?read=105063612
Here is a reasonable enthusiast from Ace's who has a reasonable statement to make about the P4EE.
And on the other hand.......
http://www.aceshardware.com/forum?read=105063634
Semi
Semi, I would have expected something more on the order of 15020%
That seems a little optimistic to me
Edit: I see you corrected that one
Keith
Thanks, and BTW, **Whew**, if it gets some people to shut the heck up around here, PLEASE........ EVERYONE....... feel free to go ahead and bookmark the 15-20% Opinion. Fine with me.
Semi
Semi, so I take it you expect Prescott to be a stellar performer, compared to Northwood, that will put A64 in its place when it is introduced, like chipguy does? Would that be your prediction?
Keith
Stellar? I don't think I would go as far as Chipguy, but those reported Process Improvements don't jibe with the reported benchmarks...... IN MY OPINION, but I don't believe the "Minor Improvement of 5-10% that appears to be in Fashion round these parts. With those types of improvements in the Process, I would have expected something more on the order of 15-20% IN MY OPINION. If that's the case, then maybe there really was some sort of serious design issue with Prescott.
(BTW, FYI, that last sentence was based that "Manufacturing Experince" Opinion, thing, for those that didn't catch it).
Sorry for the bolding, that's just for the newbies.......
Semi
And Nickel Salicide Would Be SWEET!
And the chips use strained silicon and low K CDO.
The 90 nano process includes seven layers of copper interconnect compared to six layers in the .13µ core, uses nickel rather than cobalt silicide, has 1.15u2 SRAM cells, and includes 193 nano lithography.
Dang, I did a little work with Nickel Salicide. GREAT replacement for Cobalt, but while I was working with it, there were several tough issues, controlling oxidization of the Nickel surface being the worst, as I recall, waaaaaaaaay back when. We eventually dropped it. If intel solved those issues, and incorporated Nickel, as well as strained silicon, as well as Low K Dielectric, as well as adding an additional Metal Layer, as well as additional SSE3 instructions, compared to Northwood, then......... That's quite a few Process Improvements over the Northwood P4 0.13u Process, So....... that Should lead to........
Umm, sorry, never mind, don't want to be accused of FUD.
Semi
Oh, How Stupid Of Me To Believe Intel....
And of course, who believes the Inquirer, Right?
http://www.theinquirer.net/?article=13546
But....... What if WE'RE Right.....?
BTW, I didn't see the Inq article before making any of my claims. See, it's not magic. It's just taking a little piece of data here, and a little pice of data there, and mix it up with a decade of semiconductor manufacturing experience, and apply it to the big picture. It's what PE's do every day, I can't "quantify" experience, or make anybody believe what I "THINK", but it's not insider info, or even magic, at all. Although, If I wasn't so modest (cough) What I might even say was...........
BRILLIANT!
Semi
Wow....
I use to dislike Ali Chen quite a bit, but I think I'm beginning to gain a new respect....... Just goes to show, opinions can change. Wonders never cease.
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=19678120
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=19678124
Semi
As a member of the Jewish faith, I do not take offense. Quite to the contrary, I find it a point of pride that some of Intel's advanced work is done by an Israeli development team. There are several countries in the world which do maintenance engineering, but Israel is one of the very few (2, I think) that do advanced work on new processor technology.
Israel Design should be proud of the way Banias has bloomed. I believe that might have been the first time a design didn't come out of Oregon or California. I would probably guess that there were a bunch of doubting Thomas's in those 2 other locations. Ha, So There, Take That, Oregon and Ca.
Semi
Semi, please read it more carefully. Not just the first stepping is screwed up, but the second is also no good
I think they've either gotten their steppings information mixed up, or their timetable information mixed up. Or both. Or, they're right, and I'm wrong. They say the 2nd redesign was bad, and are now having to go back and do a 3rd. I just don't believe that. No Matter, I doubt anyone was going to base their investment decisions in the next 3 weeks, on anything I say. The beginning of Feburary is the Launch, let's just wait and see. (eom)
Semi
Semi, I appreciate most of your posts,
Don't worry, Keith, there won't be any further issue between myself and mas. Thanks for the heads up.
Regards,
Semi
As it happens I did not, only now with your strong assertions
on Prescott performance/volume based it seems only on your desires and biases which shows that despite your profession you are not taking this board and its participants seriously.
Right. Desires, and biases, don't forget speculation, which is what I thought I made clear from the beginning. Pure speculation. I thought that was allowed around here. Don't worry, I won't be bothering you with my foolishness any further.
mas on ignore.
Semi
All they need is a dozen examples on Feb 2 to seed to review sites and they can worry about production afterwards.
Ohhhhhhhhhh, I understand your position now. Translation..... It's OK to Paper Launch a product, if it's an AMD product that is, but if it's an intel product....... BAD BAD intel. I get it now. Thanks.
Semi
Yes it is quite amusing if you do not want to be taken seriously .
Look mas, if you've been reading this board like you previously suggested, I would guess you've pretty much already made up your mind about me. Negatively it appears.
Works for me. You wouldn't be the first. Whatever I say..... Take it seriously, or don't, makes no difference to me.
Semi
Pleased to hear that but what 'useful' purpose does preaching the Intel 'faith' serve on an AMD Investors Board apart from pointless noise ?
Entertainment? After all, since someone keeps responding, I can't help but conclude that they at least remain amused..... Ya didn't come over here from..... where was it again......? Because it's boring here. And a person doesn't keep responding unless they're interested. What was your goal again? Clean it up?
Pretty exciting so far, Eh? You're Welcome.
Semi
---Let's assume that's true for a moment. C1 Stepping sample to vendor in December to me, means Intel would have had mid-line functionality data in mid-November on whatever C1 Stepping material was in line.(6 weeks prior, according to the sites 2-3 month production estimate noted below), at which time, they could have started the C1 Volume Wafer Starts. It's actually about 10 weeks now, but I digress. How many wafers did intel start with those samples by the way? Don't know.
Shoot, ran out of re-edit time. The reason I think that intel could have Theoretically done risk starts based on Mid Line data, without waiting for End Of line data, is because there is usually in most semiconductor manufacturing, including flash, a Post Transistor Creation Electrical Test of the Transistors, Mid-Line, before the Metal Layers are laid down. Makes sense, eh? It is also assuming that the chip needed a redesign, and the redesign that was needed was in the Transistors, not the Metal Layer. I know..... Lots of assumptions, and guesses, but I'm trying to at least make them based on some educated semiconductor manufacturing thinking.
Not much, I know.
Semi
Isn't that always the way it is with the enthusiast crowd? The unreleased, unknown product is always the most exciting.
Ha.... good point. Wanting is sometimes better than having, (Ummmmmmm....ALMOST). Actually, to be honest, I was being mean, and tryng to do a little Driod Troll Fishing (BAD Semi.... BAD BAD Semi). Guilty as charged. And there you, and Keith go spoiling my fishing fun by being polite....... Darn.
Well, I almost did catch one fish though, so I'm pretty happy.
Semi
How can you know that if you don't work for Intel ? Or is just blind faith in yout favorite mpu supplier ?
Faith, yeah, that one. Haven't you heard? the Bee's supposedly formed the Church of intel a long time ago. I should be a "Deacon" member by now.
Semi
Let me repeat again the text which predicted quite accurately what happened in December and why supply will be tight :-
---OK, point taken, I read it wrong. Although, there are several good points, there also seem to be inconsistancies......
"Intel is advancing the company internal verification of C0 step at October point in time, says that it is the schedule which ships the sample of C1 to December vis-a-vis each vendor.
---Let's assume that's true for a moment. C1 Stepping sample to vendor in December to me, means Intel would have had mid-line functionality data in mid-November on whatever C1 Stepping material was in line.(6 weeks prior, according to the sites 2-3 month production estimate noted below), at which time, they could have started the C1 Volume Wafer Starts. It's actually about 10 weeks now, but I digress. How many wafers did intel start with those samples by the way? Don't know.
- The mass production of schedule harsh Prescott
If you look at this plan, lag of μ PGA478 edition Prescott is rather heavy. Because revision change of silicon is needed, FSB 800MHz edition Prescott completely 1 means quarter to be delayed.
---Must be missing something. Volume wafer starts in mid-November means product to packaging 10 weeks later, +2 weeks packaging, That would be...... ooops, they could be right..... Sort of, My calender says no Volume until, Feb 15. Are those Prescott Motherboards ready by now..... Do ya think?
Intel is the plan that it starts the mass production with C1 step to end of year, but Qualification Sapmle (QS) with directly mass production considerably tight is schedule in December.
Maybe I'm reading this wrong, but why would intel have samples in December and not start production until the end of the year? Doesn't make sense, if I'm reading it right.
Usually, Bali day Shaun ending, letting flow the wafer to Fab, until it reaches the point where it can ship the tip/chip of the finished product abundant, there is a time lag of 2 - 3 months. Because of that, whether the Prescott tip/chip of how much quantity comes out at point in time of next year 1st quarter it is unclear.
---Q1 Quantity Unclear. Couldn't have said it better.
From the circumstance such as that, spring of next year it is the model in industry, and others there is a voice which is worried about the supply of Prescott.
---Some MoBo Makers are worried. I'd be too if intel screwed up the first stepping they promised me. Nothing unexpected about that. Intel will just hyave to give them some discounts, I guess, or maybe..... Even sell Prescott for less than P4EE. WoW, what a PR idea. As those Beer Guys would say.......
BRILLIANT!
Semi
Why not, AMD is good for another $5 after Jan 20th and Prescott ain't arriving in quantity until April/May despite the Feb 2 launch.
No, there'll be Plenty. I guess We shall see who's right....... Soon enough.
Semi
True but 10/27/2003 is the date of the report.
Right, the date of the report..... The date when this site Finally got some "information" on what they Think WAS the problem with the Prescott showed at Last Springs IDF............ and reported it. Reporting 6-9 months later, by the way doesn't impress me as a site with good insider info. But, like I said, thanks for pointing us to speculation of what WAS wrong with the first stepping........ At IDF Last Spring.
Semi
Babelfish reckons :-
"The delay reason of Prescott finally clarification
- With bus interface trouble stepping modification
The wafer of Prescott which is shown with this spring IDF.
SemiconEng reckons........ Last Spring was a long time ago. We already know the first stepping was screwed up. Why? at this point, I really don't care. I'm looking "Forward" to The beginning of Feb. But thanks for refreshing my memory.
Semi
Sounds like you have been working on it . From all the rumors it seems all that was being fixed was the heat and electrical properties. Are you saying that performance has somehow been enhanced as well, ipc wise ?
---No Not working on it, I just know intel. Been investing in them since 83'. I think it's pretty evident that based on circumstantial evidence, the first Prescott stepping was screwed up. I just don't believe that intel would delay it, redesign it, and then come out with something that was just barely as good as they originally intended in the beginning. I just don't believe it.
One of the reasons that I like intel, is, their attitude is very much like mine. And It's not what I would do as a PE. If I implement a Process Change, and it screws up, for Whatever reason, and I have to go back, and redo it, you can be DAMN sure that the second time, not only is it going to work LIKE I originally intended it to...........
It's going to work better. How? I don't know.
JMO, nothing more.
Semi
Of course, the significance of these polls is rather limited.
---My point exactly...... Thanks for picking up on that, because I notice that other unamed attempting-baiting-trollers, didn't. I knew you were smart.....er.
I doubt that many people will even notice the difference. Apart from that, Northwood is actually better suited for certain uses, as you´ll notice after Prescott´s launch.
Keith
---Ohhhhhhhh I think they'll notice a difference all right. Don't put your money down based on those 2.8GHz benchmarks you saw. Those are Ancient History. You'll see............
Semi
Prescott and Tejas although faster than Northwood may not prove so compelling against their competition, time will tell.
I guess We'll see soon enough....... If you want my advice, Don't place your bets yet.
Semi
It may sound silly, but AMD was quite restricted in Q4 by slowness of P4.
Oh? So, let me make sure that I understand the fans position...... AMD did sell a bunch of A64 chips during Q4, and even if they didn't..... It's because it's intel's fault. Thanks for clearing that up.
George Orwell would be proud, no doubt.
Semi
and also ensure they start 2004 ahead of Intel in some performance categories with reasonably priced volume parts. This should continue with the FX53 in February and s939 3700+ and FX53 in April/May.
Oh, I agree, AMD started 2004 ahead, no doubt. I also have a feeling that's all about to change. IMO, if AMD is going to succeed for The Rest of 2004 though, your "This should continue with the FX53 in February and s939 3700+ and FX53 in April/May." Prediction, had better come true..... On time. Otherwise.......
Semi
Semi, AMD Christmas is not over yet, as many hard-core and educated customers postpone the buy decision until June. There will be a second Christmas in June, maybe a bigger one. AMD must work hard to offer better performance at that moment, otherwise Intel will take the bank.
Ummmm, you're aware that Prescott and Dothan both come out in the beginning of February right? June may also be too little, too late.
Semi
Interesting poll over at Hardocp shows that the next cpu Hardocp voters (~7000) will buy in is the following popularity (rounded):-
http://www.hardocp.com/index.html?action=results&poll_ident=105
Prescott 21%
A64 3000+ 18%
A64 3400+ 14%
other 13%
XP 13%
FX51 9%
A64 3200+ 6%
Northwood 6%
Interesting to see Prescott and 3400+ so high and the 3200+ so low !
I think what's interesting is the number 1 spot is occupied by a product that isn't out yet..... And as soon as Prescott comes out, nobody will want Northwood.
Just like I said.
Semi
OT: I Hear.........
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=19670206
I Hear...... That Imitation Is The Sincerest Form Of Flattery. Thank You Sooooooooooo Much.
Semi
but what do we know about the SOI process?
Ohhhhhh Nooooo, you're not getting me to go there again
<smirk>
Semi
@Semi - A64
The hole talk seems very difficult - hopefully AMD will say
something in the coming CC.
What I saw here in germany...............
Obviously you have a better insight into the supply channel then I do. All good info, thanks.
Semi
Semi, no, A64 seems to be in tight supply in Q4, until late December, when 3000 came. It seems better now, but I still don't see a lot of A64 based computers in printed PC magazines.
OK, I'll buy that. Could be fixed now. The reason I seem to be so interested, is because during that same timeframe, was when intel's reported Prescott heat problems were happening. If A64 was delivered in quantity in Q4, then that would be pretty bad for Precscott positioning, no doubt. If not, then potentially, intel's stumble in having to do a reported restepping on Prescott doesn't seem to me to have as big an impact as I originally thought it would. Not to mention that it could indicate that AMD didn't have as merry a Christmas as they would have liked.
Just the Bee's POV..... Ya know?
Semi
Interesting Post......
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=19675621
I gotta give credit to eco, he seems to build a pretty strong case for some sort of A64 shortage during Q4, with links and everything. Granted, all the links seem to be the Inq, so take it with a grain of salt.
Now I know I'm going to get slammed for this, and sure..... maybe it's just my Bee point of view, but it seems to me, that after building this case with links, eco dismisses it with the conclusion of.......
"all just one big shift-rotated rumor fed by AMD's conservative roll out of the A64"........
---And he bases this conclusion on.......
But IIRC, around this time we began seeing A64 in ever growing numbers and only reports of mb shortages.
---So, being the open minded person that I know we all want me to be, the question I have is.... does anybody have links to verify eco's conclusion of "seeing A64 in ever growing numbers"...... during Q4?
Semi
SemiconEng, that was a good insightful post about how Intel could quietly get rid of the P4EE. I like it even better than my suggestion that they are just keeping the price point warm for an eventual Prescott P4EE!
Could be that way too.
Also, if it does turn out that P4 Northwood Desktop production gets shut off, and the flagship Desktop Microprocessor moves to 90nm, wouldn't that leave additional 130nm capacity? Let's see..... what might intel do with additional freed 130nm capacity?????
Microprocessors are intel's #1 product. Remind me again what's #2?
Semi
BTW, OK I got a little excited before, production of P4 will probably continue way past this summer........ I should have said Desktop P4. P4 Mobile and all it's variations, is another thing.
chipguy, you're missing his point. If a cheaper Prescott outperforms a P4EE, no one will buy the P4EE. So the P4EE being more expensive would seem to imply it has higher performance, no?
---Or, intel wants to drive distributors away from P4EE to Prescott, so then, they can switch off P4EE Production, and use that capacity for...... something else.
And if they deliberately want to stop selling them, why not simply stop selling them? I guess perhaps they don't want the bad PR associated with announcing a chip, shipping 50 of them 2 months later, and then announcing that they are being EOL'd.
Doug
---You don't just go into the Fab one day and throw a switch, and Ta-Da, no more P4EE. What is intel supposed to do with all those P4EE's already in the production line? Did ya consider that?
And as far as bad PR, So What. The bad PR is already out. Everyone already knows that P4EE was a stopgap intel used because they couldn't get Prescott out the door on time. It was all over the web, did you miss that? More likely, as soon as 3.4 Prescott ships, P4EE will be shut off, and in a few months, after Prescott is established, and the line is purged, nobody is going to care anymore about P4EE, if they ever did to begin with.
Looks to me, like intel is once again using the same method to drive the market away from the previous technology to the newer one that it's used before. Price pointing the newer chip below the top of the line previous chip has been done by intel a few times now, as I recall.
My guess is, that most P4 Northwood production, will all be a memory by the end of summer 2004.
Semi
Chipguy: Intel has to move into new
areas to lose money...
Uhhh, forgive me, but Intel losing money? I don't think so.
Smooth
I believe he was referring to the whole...........
"Let's move into the telecommunications business, buy buying up a bunch of small companies, that have so far, for the most part, 3 years later, drained intel resources, while not braking even, or contributing to the bottom line"......... Thing.
That loss.
Semi
1) I am still disguted with the flash memory
price hike move in the middle of a depressed
market. INTC is now in 4th place in overall
flash market share, a fall from 1st.
No doubt. I would "guess" that the person who came up with that idea, won't be receiving a big raise this year, if they're still working for intel, that is. My reaction at the time was basically WTF are they thinking?
2) It is beginning to seem that INTC has to play
catchup with 64 bit now to the desk top arena.
Here I digress. I know it sounds like intel marketing, but I don't believe, that intel believes the numbers justify 64 bit for the desktop Right Now. IMO, 64 bit for the desktop is initially going to be an adoption by the Hard Core Gamers, and a Group of "Gotta Have It's".
IMO, Gaming isn't going to be enough, Linux/Unix Near Term, and Windows Who Knows when isn't going to be enough. By the time a critical Mass of users Can and Want to migrate to 64 bits, intel Cost reduction efforts on 90nm should be steamrolling. Then you'll see IA-64 Desktop, at a price people can afford. By then x86-64 will be croaking on it's x86 compatability. You'll see.
I can only hope that the same folks who lead the flash
disaster are not at the helm of the push in the TV
chip market!!! I hope INTC will swallow some its pride
, what ever happened to the paranoid survive??? It seems the motto is now the smug and arrogant get smacked and
loose market share.
I am still very long on INTC.
Good luck to all in trading in 2004!!!!
Ronster
That decision was most likely championed by marketing. If their's one thing you can be sure of, intel won't make that mistake twice. Keep the faith, I'm sure I can see the light at the end of this tunnel.
Semi
Semi -
May I say that we all could profit from a sense of humor. Let us not take this too seriously. A laugh has investment value too you know!
I think so too.
Hey Mod person..... If I say that I wasn't offended by Doug's post, would you put it back, so I can zing him with my witty reparte'? I'll even send the post PM for preview..... Probably not, eh?
Cumon...... zinging ole SF30 is usually one of the (minor.....VERY minor) highlights of my week.
Semi