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.
Price War In Dual Core?
http://www.overclockers.com/tips00712/
From the Inquirer:
"Customers of Intel have been tipped the wink on proposed pricing of Smithfield dual core 840 (3.2GHz), 830 (3GHz), and 820 (2.8GHz) processors. . . . These chips are curently scheduled for release in the second quarter, have 2 sets of 1MB cache, a front side bus of 800MHz, and use the LGA775 socket. At launch they will be $528, $314, and $240 respectively.
The keywords there are "second quarter" and the price tags attached.
"Second quarter" means Intel will get these chips out at or before any dual-core Opterons will show up. The price tags for the latter two CPUs, especially for the 2.8GHz, is pretty shocking. $240 for a dual-core CPU is almost Celeron pricing.
Right now, AMD officially charges $418 for two dual-capable 1.6GHz Opterons and $632 for two dual-capable 1.8GHz Opterons. Intel charges the same for 2.8/3Ghz Xeons Ds.
Can you say "price war"?
AMD blames the world...
They forgot to blame the Arizona Cardinals and the price of tea in China. An oversight I suspect.
re: AMD blamed the difficulties on a severe pricing environment, weak sales in Japan, and a delay in qualifying a 256-megabit flash chip. But Handy said the pricing decline had less to do with AMD's woes than with sales in Japan and the product delay, adding that the weakness in Japan could be a symptom of the product delay.
http://zapatopi.net/afdb.html
(AMD FAN HAT ON)
No, No, No, No, No, Handy MUST BE wrong, and you've got it wrong too.... Nope, Nope, not poor Japanese sales, Nope, Nope, Not Product delays..... Nope, Nope, no contribution there, no way.... Don't you and Handy know......
When AMD is doing well, or does something right, then Hector and/or Dirk, are BRILLIANT and something approaching, if not exceeding, GODS..... GO AMD!
When AMD is doing poorly, or screws something up, it's not because AMD F***** UP.......... OHHHHHHH NOOOOOOOO, IT MUST BE..... Intel's Fault.
(AMD FAN HAT OFF)
Do ya got that "Reality" now (wink wink)? Good
Intel Keeps Ball Rolling
http://www.thestreet.com/_yahoo/tech/chriskraeuter/10204326.html?cm_ven=YAHOO&cm_cat=FREE&cm....
Looking to extend the momentum from its positive earnings announcement last week, Intel (INTC:Nasdaq - commentary - research) said Thursday that it is speeding the introduction of a new virtualization technology for desktop chips and chipsets late this year -- earlier than an anticipated launch in 2006.
On the surface, Intel's announcements and AMD's fourth-quarter problems could signal increased trouble for AMD, but the stock performances of the two companies indicate that nothing beyond the status quo is expected. Shares of AMD have held steady around a three-month low in the $15 range since the company warned on Jan. 10; Intel's stock, too, has drifted in the past week at around $22.50.
But the problem could be larger than that. Consider that the flash memory market is about to be inundated with supply, which should contribute to tumbling prices, and AMD could have missed its best, most profitable opportunity to stake out more territory against Intel.
"When that overcapacity strikes, expect NOR flash prices to collapse," he said, adding that a collapse typically entails a 60% drop in prices.
AMD blamed the difficulties on a severe pricing environment, weak sales in Japan, and a delay in qualifying a 256-megabit flash chip. But Handy said the pricing decline had less to do with AMD's woes than with sales in Japan and the product delay, adding that the weakness in Japan could be a symptom of the product delay.
Intel lets Linux into Centrino camp
http://news.zdnet.com/2100-3513_22-5542514.html
SAN FRANCISCO--Improvements to Linux have led Intel to permit Linux laptops to sport the Centrino brand for the first time.
Centrino is the chipmaker's brand name for a three-part package for mobile computers: the Pentium M processor, a supporting chipset and a wireless chip. Although Linux can run on existing Centrino notebooks, Intel until now wouldn't permit companies to sell Linux laptops using the Centrino logo.
The reason for the change: With the release of the 2.6.8 Linux kernel, the open-source operating system's power management abilities now meet Intel's requirements for Centrino notebook battery life, said Karen Regis, manager of mobile programs and promotions at the chipmaker.
ATI floods Intel Sonoma designs
http://www.theinquirer.net/?article=20809
More than 50 notebook designs
ATI captured names including Acer, Alienware, Asus, BenQ, Fujitsu-Siemens, Eurocom, Fujitsu, IBM, Gateway, HP, Medio, Lenovo, LG, Sony, NEC, Samsung, Clevo, Targa, Toshiba, ECS, Compal, Arima, MSI, FIC, Inventec, Wistron, Uniwill and Quanta and to sell them its chips for Sonoma Centrinos.
Samsung, LG Begin Selling Laptops with Intel's New Centrino Chips
http://tinyurl.com/4js42
SEOUL, Jan 19 Asia Pulse - Two South Korean personal computer manufacturers on Wednesday kicked off sales of laptop computers equipped with Intel Corp.'s new mobile computing technology, as the world's largest chipmaker launched its new Centrino family of chips globally.
Samsung Electronics Co. (KSE:005930), the country's No. 1 laptop computer maker, and second-ranked rival LG Electronics Inc. (KSE:066570) each unveiled 14 new notebook models with Intel's new chip package at a press conference in Seoul.
Dell tips hand on Intel notebook chip plan
http://news.zdnet.com/2100-1040_22-5540817.html
Dell has taken the wraps off its Inspiron 6000, one of the first notebooks to incorporate a new crop of Intel notebook chips due out on Wednesday.
The Round Rock, Texas, PC maker's Web site details the Inspiron 6000 as a 6.65-pound, consumer-oriented notebook that will feature a 15.4-inch wide-screen display. It will come with an updated version of the Intel Pentium M designed to work with the chipmaker's latest mobile chipset, Intel's Express Mobile 915 (code-named Alviso), for a price starting around $1,300 before rebates.
Intel Presler CPU slated for Q1 next year
http://www.theinquirer.net/?article=20806
Presler is a 65 nanometre dual core processor, as is its sister chip Yonah on the mobile front, and reliable sources tell the INQ that Intel already has early samples of these chips up and running in the fabs.
The firm will add EM64T 64-bit support to its existing 5XX Pentium 4s in the second quarter, and to its 3XX Celeron D processor range in the same period.
but hey I see something in the name that makes me laugh...
Yeah..... me too
I was watching ESPN Sportscenter this morning, and Linda Coen (sp?) did a story about some "Dog Award" presentation that they did at some Pro basketball game, and she reported apparently, that during the ceremony, one of the Doggies present at the ceremony, dumped...... a Turdion on the court
Intel updates Centrino notebook computer chips
http://yahoo.reuters.com/financeQuoteCompanyNewsArticle.jhtml?duid=mtfh76917_2005-01-19_08-01-46_n18....
SAN FRANCISCO, Jan 19 (Reuters) - Intel Corp. (INTC.O: Quote, Profile, Research) , the world's largest chip maker, updated its line of Centrino notebook computer chips on Wednesday with higher speeds and improved multimedia features, a move the company hopes will further propel one of its fastest-growing businesses.
The chips, which had been delayed for a few months because of a technical glitch, are expected to appear in 80 notebook models immediately, and 150 by year-end, Intel said. Pricing and other details are to be disclosed at a marketing event later on Wednesday.
Oh where, Oh where, has vapor Turion gone....
Oh where, Oh where, can it beeeeeee.....
Who knows..... Nowhere to be seen apparently
BTL
FYI, After discussion with IH Admin Matt, Andy Graves previous post was restored, due to the rest of the post being determined to be value added....... after he agreed to removal of the first sentence in the original post. Thanks.
http://www.investorshub.com/boards/read_msg.asp?message_id=5149699
If this is another bad management mistake we should certainly be pointing to it in a few years, totally right.
Smith Barney Views Reorginization as a positive
http://www-public.tu-bs.de:8080/~y0005007/intel-reorg.pdf
INTC: Announces Major Restructuring; We View as a Positive
January 18, 2005: Summary
View the reorganization from Intel as positive and in keeping with the company’s theme of "convergence"
** Sucess of Centrino platform to provide a successful model to emulate and likely provided the inspiration for the restructuring
** Elimination of ICG likely to be well received by investors, although assets remain within the company spread across newly created units
** First major move by Otellini signals his desire for change at Intel; we expect other changes are forthcoming
** No change to our estimates at this time
Sony unveils 'Centrino 2' notebook family
http://www.theregister.co.uk/2005/01/17/sony_sonoma_laptop/
Sony UK today unveiled its latest Centrino notebook, two days before the technology on which it's based will be formally launched by Intel.
That probably explains why the Japanese giant is so frugal in listing the new Vaio FS series' processor specs. However, it did let slip that the machines will used Intel's 915M chipset family, formerly best known by its codename, 'Alviso', and essentially a mobile version of the desktop Pentium 4 chipset 'Grantsdale'.
Intel to Unveil Centrino Update
http://biz.yahoo.com/ap/050117/intel_centrino_3.html
Intel to Unveil Major Upgrade of Popular Centrino Technology for Notebook Computers
SAN JOSE, Calif. (AP) -- Intel Corp. will launch a major upgrade of its popular Centrino technology for notebook computers Wednesday, as the world's largest chipmaker seeks to expand the product's audience from business users to consumers.
The new chips, code-named Sonoma, are expected to enable more consumer-friendly systems that better support video, audio and games. When Centrino was first launched, it was targeted primarily at business users.
Semi,I was merely posting information from a third party. I was interested in responses from knowledgable people that may have information about this.
Actually, you were not "merely posting information", you were responding to Elmer's response of Mas posting, and you also added your own "Opinion" here.....
Now what happens if it is not? The part probably suffers and is not packaged, the public doesn't see it (unless SI dumpster dives), and they can't determine yield issues.
That's the part of your posting I had issue with. If you had stopped at the "Now what happens if it is not"?, then that would have been seeking information. However, when you continued on to answer your own question, with "The part probably suffers,and is not packaged.... etc", then IMO, you crossed over into personally agreeing with the same FUD the Author was feeding people. I don't call that "merely posting information". In any case, I agree, in the future, I won't try and tell you how Chip Distribution works, because I don't know a damn thing about that, If you do the same, when it comes to making chips.
HP to rejuvenate OpenVMS on Monday
http://news.zdnet.com/2100-3513_22-5537046.html
Hewlett-Packard plans to inject some new life--or more accurately, a new processor--into its venerable OpenVMS operating system on Monday, CNET News.com has learned.
The Palo Alto, Calif.-based computer and printer company is launching version 8.2 of the operating system, which for the first time will bring OpenVMS to Itanium processors, sources familiar with the plan said. The announcement will accompany a refreshed Itanium server product line that HP will trumpet during a Webcast with Chief Executive Carly Fiorina on Tuesday.
Leading Online Marketplace Initiates DDR2 SDRAM Trades.
http://www.xbitlabs.com/news/memory/display/20050117105125.html
One of the premier online exchanges for dynamic random access memory, DRAMeXchange, initiated Monday coverage of spot-prices for DDR2 SDRAM, which may indicate that the new type of memory is becoming more and more popular on the market place.
Elmer, any chance that IBM could shift resources over to 65nm without mastering 90nm first? I don't know if there are any dependencies at 90nm that are necessary before a manufacturer attempts the next node, but if IBM can shift resources and get 65nm working by mid next year, isn't that better than trying to improve yields on 90nm and having 65nm slip until 2007?
Whew, this answer might be a little long, so bear with me
Without knowing specifically what portion of the Process is IBM's problem, it's difficult to say whether jumping to 65nm would help. From what I've seen, The way the Process Development works, is that The Development PE's work on their specific area of the process, then "Integrate" the whole Process together. The Integration part is where the major probelems normally crop up.
Let's say for example, that there is this Etch Process Engineer, dveloping his section of the Process. His Primary task, is to get his etcher to etch the specific material required, without etching the materials that he doesn't want etched, which involves trying different chemistries vs. Power Levels vs. etch times to remove the specific materials at the required rates, without removing any materials that he doesn't want to.
Even after all that work, then he has to integrate it with the upstream Engineer, who is working on his Material Deposition Tool Process, and downstream with the Engineer who is working on his Wet Etch Portion of the Process, and get all that working, then, there may be unforseen issues even 3 or 4 Steps downstream in the Process, or even several layers above or below the specific layer the Etch Engineer is working on.
Even after all that (thought I was done eh?) frequently, what works just fine and dandy at low volume, can often blow up in your face when you try to ramp the Process from low volume to high volume.
I BELIEVE that this "Ramp It Up" issue (whatever it specifically is) is the part IBM is having difficulty with "Now". I think they initially had big issues with the integration part itself, then they got that fixed, and they stupidly started talking to reporters about how great things were working, then they tried to ramp it up, and it blew up in their face.
If it's something simple, like a "Specific Tool" not being able to keep up, or perform as expected at High Volume, then maybe you can dump that section of the process by going to 65nm. If it's some sort of "Integration Interaction" that is not fully understood, then moving the Process to 65nm might make it worse. It also might make it better, but that's a BIG Bet to make with your 65nm Process Dollars. From what I've seen, most Management Types would prefer to fully understand what's wrong at 90nm, and fix it before attempting 65nm.
BTW, from my experience, the ramp issues are usually of the "Integration Interaction" variety, not the "Specific Tool" variety. From the extensive amount of time that it appears to be taking IBM to fix 90nm, I'm GUESSING that whatever the specific problem is, it's also of the integration variety. Good Luck IBM........ NOT
JMO
(contrary to popular belief I never wear a droid hat so I can't take one off like you just did :)
I stole it from "Doug". Looks like he got a replacement though
http://zapatopi.net/afdb.html
but you are not addressing the faulty logic which was really the point I was making..... but I do know faulty reasoning when I see it. To see a perfectly controlled process step and then conclude that perfection indicates a problem is the type of logic that I thought was the sole province of the AMD Fanboys. Your data now makes it clear that perfection is indeed a sign of perfection! Who would have thought?
Yeah, I thought you were already doing a good job of that, here and on SI, and as far as Droidian Logic..... I guess I forgot about that.... Wait a minute, here ya go..... (Droid Hat On) Hey, don't you know? when AMD does something perfectly, then Hector and Dirk are Gods, and it's destined and planned to be that way, GO AMD! but if intel does something perfectly, then they obviously screwed up, and there must be something wrong with it somewhere, or they twisted somebody's arm to do that, or they gave special pricing to buy that, or they paid off the writer to say that, cause it can't be that they did something right, or better then AMD or anybody else. So there..... (Droid Hat Off).
Whew.... that "AMDroid Logic" nearly burned my neurons.... Good thing I got that Hat off in time
Elmer,
I don't think they assumed Intel had a lower yield, just that
all of the data they have examined indicated that the alignment was perfect. Now what happens if it is not? The part probably suffers and is not packaged, the public doesn't see it (unless SI dumpster dives), and they can't determine yield issues.
That is not correct. They did not "Examine any data". They viewed Scanning Electron Microscope Images, and drew a conclusion based on those images, and attempted to suport their position by using vague "slight misalignment" (ie: No Data) wording. The contention that misalignment of the contacts leads to yield degredation is an incorrect assumption based on outdated Process Technolgy in use prior to 130nm. On the other hand, I DID use data, as I have already pointed out here....
http://www.investorshub.com/boards/read_msg.asp?message_id=5110023
To prove that their "conclusion" is FALSE. I pointed out a link that specifically gives exact misalignment tolerances, on Unlanded Contacts, and a second link that specifically points out that intel has been using unlanded contacts since the 130nm process, instead of using the vague "slight misalignment" B.S. conclusion stated in the article.
I have emailed my posted link to the author, who so far has declined to respond. Since the data numbers in the article on unlanded contacts, refutes their "contact misalignment leads to lower yields" statement and since the second link I provided states that intel has been using unlanded contacts since 130nm, this Data specifically refutes their "conclusion", so I would appreciate it if you would please do not continue to spread FUD, about things you know nothing about. Thank You.
An amazing conclusion from that article:
SI has determined that the Intel devices are most likely very sensitive to slight mask misalignments.
Baloney. This is an incorrect assumption, based on old Process Technology. It's too bad that the Author either does not know what he is talking about, or else, did not do his homework, to investigate that since the advent of "Single Damascene" Process, the now common usage of Unlanded Contacts, including intel's use of a Single Damascene Process" for the Metal 1 Layer, since the introduction of 130nm, allows a misalingnment tolerance of the vias of 0.03 microns (30 nanometers), which makes his statement of
The fact that every transistor contact was perfectly aligned in the Intel device tells us that the yield was highly sensitive to slight alignment problems and that wafers with chips exhibiting such problems failed wafer test. Such sensitivity will reduce the wafer yield considerably below the wafer yield achieved by TI.
Complete, Total, and Utter Bullsh*t.
Single Damascene Process Information
http://www.cyberfab.net/icnotes/icnotes9.html
Interconnects:
ILD is FSG with a dielectric constant of 3.6
W plugs are used to contact the silicided regions. Unlanded contacts use Si3N4 as etch stop layer.
Interconnect architecture and performance:
To facilitate the use of local routing, medium interconnect and power bus routing, three Cu thickness are used for the 8 levels; the target sheet resistance for the 3 thicknesses are 85,63 and 24 mW/square.
The stacked via resistance can tolerate a misalignemnt of up to 0.03 µm. The ILD is FSG for low cost-performance ASICs while two other low k materials were investigated for high performance.Going from FSG to low k materials (low k1 and low k2, no further precision) allows a 22 % and 29 % improvement in RC delay while the total capacitance difference between the two low k materials is only 3 %. However thermal/mechanical, manufacturability and reliability are very different and may rule out one material in favor of the other. All metal levels support J max > 1E6 A/cm2 at 110°C.
Intel's Single Damascene Process Info
http://www.intel.com/technology/itj/2002/volume06issue02/art01_130nmlogic/p05_interconnects.htm
INTERCONNECTS
Chip performance is increasingly limited by the RC delay of the interconnect as the transistor delay progressively decreases while the narrower lines and space actually increase the delay associated with interconnects. Using copper interconnects helps reduce this effect. This process technology uses dual damascene copper to reduce the resistances of the interconnects. Fluorinated SiO2 (FSG) is used as an inter-level dielectric (ILD) to reduce the dielectric constant; the dielectric constant k is measured to be 3.6. Figure 17 is a cross-section Scanning Electron Micrograph (SEM) image showing the dual damascene interconnects.
Table 1 lists the metal pitches. The pitch is 350nm at the first metal layer and increases to 1200nm at the top layer. Metal aspect ratios are optimized for minimum RC delay and range from 1.6 to 2. The first metal layer uses a single damascene process, and tungsten plugs are used as contacts to the silicided regions on the silicon and poly-silicon. Unlanded contacts are supported by using an Si3N4 layer for a contact etch stop. Copper interconnects are used because of the material's lower resistivity.
Ya know, and I'm just noodlin' here, but, if Intel did "OK" after spending 27 Billion on capacity and all of their major competitors suck, that could mean that Intel is taking market share from everybody.
Did Samsung report yet on it's Flash memory sales for Q4? Just curious
Three warnings show chip sector faces tough 2005
http://tinyurl.com/6fpvs
Europe's biggest chip maker, Infineon (IFXGn.DE: Quote, Profile, Research) of Germany, and second-placed French-Italian firm STMicroelectronics (STM.PA: Quote, Profile, Research) both said earnings would fall significantly below market forecasts. Worldwide in 2004, Infineon ranked number four and STMicro number seven. In the United States, the world's biggest chip maker Intel (INTC.O: Quote, Profile, Research) beat expectations this week, but much smaller rival AMD (AMD.N: Quote, Profile, Research) missed Wall Street estimates by a wide margin.
The only reason that I can think of is...... It must be intel's fault (tm: Droids)
Of course it is. Intel released Nocona 6 months ago and
it had the fastest ramp ever for a Xeon product and out
sold Opteron by a country mile.
HA.... I agree. I was hoping that "someone" would pick up on that irony. Although you are not that "someone", Good job catching that CG
ATI: PCIe-enabled notebooks to account for 60-70% of global shipments in summer
http://www.digitimes.com/bits_chips/a20050113A7037.html
About 60-70% of global notebook shipments this summer will feature PCI Express (PCIe)-enabled graphics solutions, buoyed by the launch of Intel’s Sonoma platform later this month and the support of PCIe-compliant graphics chips by the leading notebook vendors, stated Rick Bergman, senior vice president and general manager of desktop business unit at ATI Technologies, yesterday in Taipei.
Euro AMD Opteron server demand slows
http://www.theregister.co.uk/2005/01/13/amd_euro_share_slide/
Sales of AMD Opteron-based servers in Europe appear to be slowing, market watcher Context warns. After tracking Opteron sales throughout Europe's seven biggest economies, Context today said that sales growth collapsed during the early month of Q4 2004.
The only reason that I can think of is...... It must be intel's fault (tm: Droids)
WinXP x64 Not Meeting The Hype?
http://www.gamepc.com/labs/view_content.asp?id=xp64rc1&page=9&cookie%5Ftest=1
Performance wise, the only people who are going to benefit from x64 Edition right off the bat are high-end workstation and server users who will benefit from the expanded memory capabilities of this OS. Desktop users will see no performance gains when using traditional 32-bit applications. However, IF AND WHEN more software developers choose to release true 64-bit executables, it’s certainly POSSIBLE that 64-bit systems will show performance gains over 32-bit systems. As of right now, with basically zero performance gains over 32-bit, it seems unlikely that there will be a great shift to 64-bit when this software is first released. However, IF for example, popular games like Half-Life 2 or UT2004 show significant performance increases when 64-bit executables are released, the move towards 64-bit Windows would quickly pick up steam. As of right now, there is no shining example out there showing that running a 64-bit OS is any better than a 32-bit OS.
Tech Report on 90nm Opteron delay
http://techreport.com/etc/2005q1/ces/index.x?pg=1
AMD said that 90nm Opterons were not shipping yet, but will be in the "very near future."
Yeah, sure. So much for IBM/AMD's supposed superior 90nm SOI process
Intel’s Unannounced Pentium M Chips Sneak into Retail
http://www.xbitlabs.com/news/mobile/display/20050111091727.html
Intel Pentium M with 533MHz Processor System Bus for Sale in Japan
Yet unannounced Intel Pentium M processors for 533MHz processor system bus emerged for sale in Tokyo, Japan. The products come in Intel’s typical retail packages, which implies that the chipmaker has already started to ship such central processing units to various customers including retailers.
Having Retail boxed processors for sale BEFORE the official announcement, IMO bodes very well for Yields and Bin Splits for this processor, while AMD's "Turdion" Might be released sometime in H105..... maybe
The press sucks
Their headlines read:
"Intel reports record sales, but earnings fall"
Oh, I don't mind the few naysayers in the Press (and elsewhere), most companies would kill for 56% profit margin, including AMD. It sounds like Sour Grapes to me, and besides, overall, those people are more then offset by the positive Press Headlines....
http://tinyurl.com/695xl
Intel posts record revenues
http://www.eet.com/semi/news/showArticle.jhtml?articleId=57700593&kc=2515
Intel reports record sales, will boost 2005 cap ex
http://tinyurl.com/57o4y
Intel gains on strong Q4, outlook
Inventories cut nearly 19 percent; Q1 guidance in line
http://biz.yahoo.com/rb/050112/tech_intel_stocks_1.html
Intel Shares Rise After Quarterly Results
"Turion: A New Name For An Old Promise?"
What's A Turion? We explain that outside the name, this is nothing new for AMD, and that in fact, the talk about it is just a delay in disguise.
http://www.overclockers.com/tips00708/
There was some hooping and hollering last week about this new AMD mobile processor line called "Turion." I fail to see what the fuss is all about, given what AMD has said in the past about it. Yes, Virginia, AMD has talked about this in the past.
P.S. Per this, the name "Turion" itself is supposed to evoke connotations with the word "tour." Personally, it makes me think of "centurion" when I'm feeling kindly about the product and "turd" when I'm not.
The AMD Turd..... I like it
S.Africa's Telkom signs memo with Intel on WiMAX
http://yahoo.reuters.com/financeQuoteCompanyNewsArticle.jhtml?duid=mtfh43289_2005-01-11_09-42-29_l11...
JOHANNESBURG, Jan 11 (Reuters) - South Africa's Telkom (TKGJ.J: Quote, Profile, Research) has signed a deal with Intel (INTC.O: Quote, Profile, Research) to test WiMAX, the latest long-range wireless technology, the state-controlled phone company said on Tuesday.
Cnn Headline News.....
Reporting on AMD's warning, just called them...... "That memory Chip Maker"
Yes, the same products are built using 180, 130, and 90nm, but Intel started the 90nm conversion for flash a few quarters ago. I suspect a large portion (50%) is now on 90nm... However, they have not yet introduced strataflash on 90nm... it is still sampling.
hope this clears things up a bit,
Alan
Thanks.
Some thoughts on flash...
Intel is now in volume production of 90nm NOR, while AMD is split between 130nm and 110nm... cost advantage to Intel for "standard NOR" products.
While I agree with most of what you say, I don't believe that with 3 Operating 90nm Fabs, that intel is in "volume production" of NOR on it's cutting edge 90nm Process. I would say that the majority of intel NOR is actually not on 90nm, but rather most of the volume is on 130nm. Intel's own web page states that NOR is being produced on 90nm, 130nm AND 180nm. Since 180nm and 130nm fabs are mostly paid for, I agree that cost advantage is to intel, but I don't believe that intel would use it's cutting edge process in "volume" for a low margin product like Flash..... IMO.
http://www.intel.com/design/flash/intro/90nm.htm
Intel® Wireless Flash memory is now available on industry-leading 90 nanometer process lithography, as well as existing 180nm and 130nm. Intel NOR flash memory and 90nm lithography have several highlights:
Keith tries to moderate their
exuberance into realistic optimism but he is regularly
boo'd down by the bipolars on the upswing.
I know, I admit, I initially mis-judged Keith as a Droid. My bad, and my apologies to Keith (I know you read this thread Keith :))
OK OK, it's also partly becuse I noticed he took my name off of his ignore list in his profile
It will be fascinating to see if mas_ will eat crow after Intel's announcement tomorrow. Of course, his AMDroidian conditioning will prevent him from remembering that he made his SI statement in the first place.
Yeah..... but I have to give ole mas a little credit for at least seeing the biggest "Droid" for what he really is, and I admit, that so far, the best part is seeing ole Doug slapped around, on the SI AMD thread, no less. It almost makes me like Rupert and mas, and sort of makes all the crap I took from Doug/UPC/Conductor almost worth it. What goes around, comes around, eh?
http://www.siliconinvestor.com/readmsg.aspx?msgid=20936574
http://www.siliconinvestor.com/readreplies.aspx?msgid=20936574
Thanks for that article, it was interesting... the price we pay to keep our jobs.
thanks again.
Yeah, don't get me wrong, they put you up in a VERY nice apartment, pay for your utilities, cable, phone, get you a rental car, pay for your food, provide regular trips back where you came from, etc etc etc....
But basically, you're there in a strange place, you're usually unfamiliar with the area, you mostly have no friends beyond the other people you're on assignment there with, it's hard on your family, you're sleeping in someone else's bed, living in someone else's place, and usually working long long hours.
The novelty wears off quick. Trust me
AMD warns revenue to fall below estimates
Ha..... I guess DRBES was right for a change, AMD's earnings certainly are going to "stun" some people, all right
http://www.siliconinvestor.com/readmsg.aspx?msgid=20929665