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.
Is AMD still having execution problems?
http://www.theinquirer.net/?article=6081
Expect no Athlon 2600+ chips for weeks yet
(Sarcasm On)
Whatever could you POSSIBLY mean Fin??? Why, you're not suggesting that "Gimme Mine" Jerry Sanders was LYING when he said this back in late April 2002 are you??
http://zdnet.com.com/2100-1103-915032.html
"Days later, AMD Chairman Jerry Sanders gleefully responded to Intel's claim of superior yields.
"How do you spell 'bull****'?" Sanders asked investors and analysts at Merrill Lynch's Hardware Heaven conference in San Francisco in late April. "The only way they could do that would be to invent a perpetual-motion machine...We will put our yields up against anybody's."
Why, it was only a few months later that the 2600+ was "introduced" (cough, cough) and it's ONLY been 2.5 MONTHS since AMD "Introduced" (Cough, Cough) the 2600+.
http://www.theinquirer.net/?article=5066
Naaaaaaawwwwww..... No problems there......
(Sarcasm Off)
Semi
Subzero, I think Intel has already proven with their last few releases that they only intend to have volume launches from now on.
I have a feeling that when the 3GHz launches in 9 days, that there will be plenty of chips available for OEM's, but I don't think that you'll see them on Pricewatch...... for a week or so...
Semi
The prices on these handheld systems is very aggressive. And very tempting, too. I think Dell's goal of 25% market share is possible. I also see that they plan on going after Cisco in the networking market. It might be a good idea to invest back in Dell.
wbmw
Is this really a big deal though?? I don't know, maybe it's just me, but about a year ago, I saw everybody around at work runnding around with PDA's, but the novelty seems to have work off a bit. The whole PDA Market seems to be a bit depressed, and I don't see quite so many people using them these days. Maybe just my perception thing though.....?
Semi
So this must be step one in AMD's cost cutting plan. If true, it is as I expected. The easiest way to put a dent in the bottom line is to cut staff. The first question is, who has been cut? The second one is, what is step two?
wbmw
If true, the most logical place to cut, is to take a chunk of the Fab workers. "Operators" comprise a fairly large portion of the staff, and if things do turn around, they can be rehired, or new people can be trained to run the tools in the shortest amount of time. Bringing Process, Yield, Integration, Device Engineering etc up to speed takes much longer.
The Training Department is also a good place to cut back. I've had several offers to move over to Training, but from what I've seen, when things go bad, the Training department is the first to go.
Step 2, could be some sort of Warm down of the Factory for brief periods. I believe intel previously forced Ireland Employees to take "Time Off" without pay, or use up vacation. If there are any Production lines that are not running at capacity, it might be wise for them to take some time off. That way, AMD can save some money on thew books by not running 1 or both Fabs over the holidays.
Might also be time to Dump Alchemy (sp?)
Semi
Kind of interesting...
Apple going Intel....
http://story.news.yahoo.com/news?tmpl=story&ncid=77&e=4&cid=77&u=/mc/20021028/tc_mc/....
Report: Apple will go Intel next year
Mon Oct 28, 2:03 PM ET
by Dennis Sellers, dsellers@maccentral.com
Apple's "continued technical disadvantage" -- which we assume means the race for computer processing speed -- against Intel is expected to force it to adopt x86 technology by the end of 2003, according to a new report by Giga Information Group Inc., a global technology advisory firm.
--- Actually, it's Yahoo that says intel. what the report says is x86 technology. That's not necessarily intel.
Semi
re: I don't want AMD to go out of business.
Employees and stockholders don't always have the same interests. I would like to see the margins increase, and the stock price go up.
John
---Sorry, I'm not not an employee of AMD, and not an AMD stockholder either. I'm not in favor of AMD's margins increasing nor having their stock price go up. You must again be referring to someone else. Please see my "Constrictor" reference in previous post.........
If intel continues the slow constriction of AMD, without killing them, for far into the foreseeable future, than that's the best thing I can think of.
Semi
Semi
I'm pretty glad that at least someone is asking me about MPF. I was beginning to think I was a wasted resource here.
wbmw
No, not wasted, I just had a rough week, that's why I didn't get to read your previous MPF stuff, and it was taking me awhile to get through the "NOISE"..... Ya Know....??
Thanks For the Update.....
Semi
On another topic, how 'bout that AMD stock? The market seems to disagree with our thesis that they are going out of business.
John
Our Thesis???? Not mine Baby..... I don't want AMD to go out of business. Competition is good. Intel got fat and lazy without competition, and I for one have no desire to go back to those days.
As far as the Market is concerned, I think the market is responding right now to the hype. The thing about hype, is that it all looks like great fun for awhile, but eventually you have to deliver. And that from AMD, I don't see.
Many Fanboys in the past have crowed about how Pricewatch is the measure of the market. They ripped intel's lack of pricewatch listings if they didn't show up on the same day as introduction. Of course now that the shoe is on the other foot, the story runs more like "I saw 3 - 2400+ listings on Price Watch only 2 months after introduction, WOW that's a good thing".
No, I don't believe that AMD has their 0.13u issues worked out yet, and I'm not interested in the death of AMD. No, it's much better to keep them around to keep the DOJ away, and keep intel sharp. If intel continues the slow constriction of AMD, without killing them, for far into the foreseeable future, than that's the best thing I can think of.
Semi
Tim -
Rather than annoy people here I have taken this discussion to the "Parking Lot".
We can continue it there and anyone who gives a crap can follow.
EP
Good..... Go.... Both Of You.....
Semi
In these images, you see Banias running Photoshop, and a probe is attached to measure power dissipation. Clearly, Banias is going to be a cool chip, with power at 0.25-1.30W. Wow.
Thought that some people might be interested.
wbmw
Well I am. Finally, something intel related..... I haven't been reading the MPF stuff, you posted the programs that Banias was running, did you see anything running on Hammer?? Was it a static display??? Did they release any frequencies??
Semi
Huh? Willamette was the first generation P4.
Woooops...... Too early this morning...... I meant Merced. Intel is still running their 0.18u Fabs as far as I know, and with most of the cost reduction done, couldn't they just move that design over to 0.13u, like they did with P4 and Celerons, and continue to crank those out and sell them as desktops???
Semi
Semi,
I just a bit concerned about this part:
Analysts said that, although the disparate processor architectures won't necessarily compete for the same sockets, the fact that AMD and IBM are following similar paths could apply pressure on Intel Corp.,...
You know what the old generals say about running a two front war. Will Intel need to have 64/32 bit on the desktop to compete? Is their decision that the desktop doesn't need 64-bit a longish term strategic mistake?
Time will tell.
John
Analysts say lots of things, and the phrase "could", doesn't necessarily mean "will". I recall Analysts saying that Transmeta's Caruso "could" destroy intel in low end/low power. I'm still waiting....
Also, I don't recall intel saying that 64 bit wasn't needed for the desktop longish term. I recall them saying it wasn't needed for desktop right now, and they seem to be right. By the time IBM gets their first generation 64 bit going, Intel will be full steam into 3rd generation IA-64 Madison, and most likely, Willamette, will be transitioned to 0.13u. How cheap do you think a 0.13u IA-64 Willamette will be by then?? Cheap enough for desktops??? It "could" be.
What I notice, is that all of those "could" scenerios, seem to assume that intel will be standing still with IA-64 in the meantime. They're not. I agree, time will tell.
Semi
Gala said the new processor will be aimed at desktop PCs and workstations, entry-level servers, electronic games, networking equipment, storage, and high-end set-top boxes. One major potential application is in a new server line developed by PowerPC user Apple Computer Inc. Tom Halfhill, an analyst at In-Stat/MDR, San Jose, said IBM will likely use the new processor to win more sockets at Apple, which has historically bought most of its PowerPC chips from Motorola Inc.
Though neither IBM nor Apple would comment, the PowerPC 970 is equipped to run the Apple Altivec multimedia instruction set for servers.
It doesn't sound too good for Motorola. On the other hand, First Silicon is one thing, manufacturing it in volume is something else. I'm impressed with IBM's capabilities in making very good chips in small volumes. I'm not convinced of their abilities in manufacturing the higher volumes that their "aims" would require. It remains to be seen whether it will offer competition for IA-64.
Semi
AMD To Be Pushed Down To # 3???
http://www.theinquirer.net/?article=5814
ST Micro "in talks" to buy Motorola chip operations
Deal to create the world's second-largest chip company
By INQUIRER staff: Wednesday 16 October 2002, 11:31
ACCORDING TO A REPORT in the Financial Times, ST Microelectronics is in talks with Motorola to buy out its chip operations.
Maybe AMD should be worrying more about STM/Moto than intel???
Semi
I think Bryant's answer was that they were expecting to get even more cost out in the quarter but it didn't happen.
Bryant said you could shut a line and take a one time write off for the remaining depreciation... but they prefer to risk underutilization in order to have capacity available for an up tick (the investor view versus the trader). The good news is the capacity can still be used for saleable products. This puts Intel in a very good position for an upturn, but they do have to deal with the potential for underutilization charges each quarter.
This seems to verify what I said before regarding startup costs vs. cost reduction. When starting up the 0.13u process, intel probably cared little about the cost, and once things were up and running, then the cost reduction stuff probably started. Maybe intel didn't start the cost reductions soon enough for it to pay off this quarter, combined with the continuing sluggish sales, resulted in less payback then expected?
As far as the underutilization, if there was some sort of cutback in wafer starts, then the installed tools could be turned off temporarily. In that way, even though you would still be charged some amount for the capital expense depreciation, you could still save some money on Maintanence Labor Costs, Spare Parts, Electricity, etc. Not much savings, but something. Then, as you say, when things get better, then you turn the tools back on.
Semi
Had you heard the name "Anvil" before (I'm assuming they are referring to what I think has been called Yarnhill). Anyway, it's kinda cute. What's the one thing that can stand up to a "hammer"? LOL
I also agree with the article: I don't think a product such as "Anvil" will see the light of day (but the possibility sure does keep AMD off balance, doesn't it?)
greg.
Nope, Haven't heard that name mentioned. BTW, I Haven't heard anything said to me about a Yamhill either. Sorry.
Semi
In AMD's case, the analysts don't believe the Opteron meets any of the criteria for new processor adoption--application vendor enthusiasm, server vendor commitment, server vendor enthusiasm, or and customer enthusiasm. Even with OS support from Windows and Linux, unless there's a change in those critical aspects of Opteron's ecosystem, I am standing by my previous assessment: that when it comes to enterprise computing, Opteron is going to be a dud.
Opteron being called a "dud". Kind of ironic.... eh?
Semi
wbmw, Dual 1.8Ghz 64-bit cores? Sounds cool and it may stabilize Apple's sales. I guess there go the dreams of many AMD fans to have Apple go with Hammer... On the other hand this shows that more companies are trying to find ways to improve cpu throughput without going for Max Mhz/Ghz.
C
I could have sworn I heard some rumors that intel was looking at going with some sort of dual core chip. Don't remember hearing anything about them actually getting it working. Maybe Moto will have more success. We'll see.
Semi
Next week's conference calls should shed a lot more light on things, hopefully.
wbmw
I think recent reports from motherboard makers of increased sales also bodes well for some sort of recovery going forward. How strong it will be, seems uncertain right now. I'm keeping my fingers crossed......
Semi
Should P.E.'s Habla Espanol???
http://www.siliconstrategies.com/story/OEG20021011S0046
Como Esta Usted??
Semi
Nikon Files Additional Litho Stepper Lawsuit Against ASML
http://www.siliconstrategies.com/story/OEG20021011S0025
TOKYO -- Increasing its legal assault against its rival, wafer stepper maker Nikon Corp. filed patent-infringement complaints against ASML Holding BV (Veldhoven, Netherlands) in Japan and North Korea this week, following moves initiated last year in U.S. courts and with the U.S. International Trade Commission.
I wonder if this is going to hurt 193nm Stepper Deliveries, and thus 0.09u Development/Implementation???
Semi
Semi, thanks, I wanted to check up on you guys since the SI intc thread has been overrun by all the intc shorts. To get back on topic, I think that AMD & Intc will do well in '03 unless we have a double-dip recession. All the computers bought in the late 90's will be breaking down soon and fixing them will be more expensive than purchasing new ones.
C
I think that intel is even better positioned for recovery going forward, then most believe. Why? Cost Redction. This is speculation of course, but based on previous start-up I've done, when you start up a new Process, you pretty much ignore cost. All the PE's care about is getting the durned thing working. Nobody even looks at cost.
Once you get the Process stable, and yields are up to expectations, the PE's job shifts a little to Sustaining, Cost Reduction, and Process Improvement. From what I've seen.... in that order.
It appears that intel's 0.13u Proces is now "stable", so I would guess that major cost reduction focus is starting, if not already started. I expect significant cost savings for the next few quarters at least.
Since AMD is still seems to be wringing whatever speed is left out of 0.18u, I doubt there is much cost reduction savings available on that Process anymore, and If AMD follows the same "just get it working" strategy with their 0.13u (and they better be), judging on the results so far, I don't believe they can do much cost reduction in the next few quarters unless they get their 0.13u Process stable.
All speculation of course, JMO.
Semi
John, you and wbmw must be having funky (I don't want to use the other word ) dreams about AMD going out of business.
Charles Gryba....... Welcome to iHub!!!!
(You're signature gave it away)
I don't want AMD to go out of business. So far, AMD has been very successful..... in keeping the DOJ off of intel's back. The only thing I really miss, is Jerry Sanders proclaiming that AMD is taking share away from intel. It was the best shield intel had...... Sigh, oh well, maybe Hector will be more vocal in the future.
Semi
John, with $3B in sales with the joint flash venture and no CPU losses holding them down, AMD could once again become a profitable company. Why let the competitive CPU market weigh them down?
One reason...... Jerry Sanders. His hatred of intel, and his presence still on the Board Of Directors. It might give people the idea that his Jihad against intel failed, and I don't believe he will let that happen.
Semi
Elmer, looks like AMD's 2400+ CPUs are still MIA! I guess you can just about count October as a non-revenue month for AMD in terms of CPUs.
Overclockers.com is tracking availability of the 2400+ on Pricewatch, so far..... MIA
http://www.overclockers.com/tips00138/
Accubyte No Pricewatch blurb indicates ETA 10/29/02
buymstar No Only real indicator is product webpage where they state "limited quantity on 1st shipment."
cbwnet No Product webpage indicates ETA 10/28/02
PCProgress No Product webpage indicates ETA 10/14/02
Monarch Computer No Per webpage, orders made by 10/6 will start shipping 10/11; post 10/6 orders will start shipping 10/17.
Lynn Computer Products No ETA mid-October per Pricewatch comments
Spartan Technologies No Product webpage indicates ETA 10/14/02
Computerhq No Webpage only indicates "due in"
Newegg No 2400+ has briefly been up for sale twice over the last week, then yanked. No Pricewatch price recorded.
Semi
http://list.realestate.yahoo.com/re/homevalues/
That's it!!! I saw that with a Google search, but yesturday I couldn't get to it. Maybe the server was down. Thanks.
Semi
OT: House Valuations
Hey, Sorry about the off topic, but I used to have a weblink for finding out what your house is worth. Now I can't find it. Does anybody know of the site where you can plug in your house location, and a few other pieces of data, and get back a "guess" on it's value??
Semi
Intel to Announce Chip Sets Supporting Hyper-Threading
---And Look At All The Boards Available..... Already.
http://www.xbitlabs.com/news/
Bunch of i845GE and i845PE Based Mainboards in Japanese Retail [11:07 am] Anton
Tomorrow Intel will officially announce its i845PE and i845GE chipsets with PC2700 memory and the Hyper-Threading technology support. As always, the land of the Rising Sun is ahead of the whole world: already yesterday a lot of Japanese companies started to sell mainboards from different vendors in retail. The prices on the novelties vary greatly: from affordable $100 to expensive $175.
Here is the list of the i845PE and i845GE powered mainboards with brief specifications that made their appearance this weekend in Japan:
Gigabyte GA-8PE667 Pro: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 1 CNR; 2-channel ATA-33/66/100; 10/100Mb/s Ethernet controller; 6-channel AC'97 sound; USB 2.0; Dual BIOS, etc; ATX. $110-$120.
Gigabyte GA-8PE667 Ultra: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5PCI; 1CNR; 2-channel ATA-33/66/100; ATA-33/66/100/133 RAID controller PDC20276; 10/100Mb/s Ethernet controller; 6-channel AC'97 sound; USB 2.0; Dual BIOS, etc; ATX. $135-$145. For more details see this news-story.
ASUS P4PE/LAN: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 6 PCI (including BlueMagic slot); 2-channel ATA-33/66/100; Broadcom 10/100Mb/s Ethernet controller BCM4401 KFB; 6-channel AC'97 sound; USB 2.0; etc; ATX. $115-$130. For more details see this news-story.
ASUS P4PE/GBL/SATA/1394: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 6 PCI (including BlueMagic slot); 2-channel ATA-33/66/100; Promise Serial ATA-150 + Parallel ATA-33/66/100/133 RAID controller PDC20376; 10/100/1000Mb/s Ethernet controller Broadcom BCM5702 CKFB; 6-channel AC'97 sound; USB 2.0; FireWire (IEEE-1394) controller VIA VT6307; ATX. $160-$175. For more details see this news-story.
ASUS P4GE-V/LAN: i845GE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; Integrated Intel extreme Graphics core; AGP 4x, 6 PCI (including BlueMagic slot); 2-channel ATA-33/66/100; Broadcom 10/100Mb/s Ethernet controller BCM4401 KFB; AC'97 sound; USB 2.0; etc; ATX. $127-$130. For more details see this news-story.
MSI 845PE Max-L: i845PE; 2 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 1 CNR; 2-channel ATA-33/66/100; 10/100Mb/s Ethernet controller; 6-channel AC'97 sound; USB 2.0; etc; ATX. $115-$130.
MSI 845PE Max2-FIR: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 1 CNR; 2-channel ATA-33/66/100; Promise Parallel ATA-33/66/100/133 RAID controller PDC20276; Intel 10/100/1000Mb/s Ethernet controller RC82540EM; Cmedia CMI8738MX 6-channel sound; USB 2.0; FireWire (IEEE-1394) controller VIA VT6306; ATX. $145-$155.
ABIT BE7: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 2-channel ATA-33/66/100; Realtek 10/100Mb/s Ethernet controller RTL8100B; 6-channel AC'97 sound; USB 2.0; etc; ATX. $130-$140. For more information see this and this news-stories.
ABIT BE7-RAID: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 2-channel ATA-33/66/100; High Point ATA-33/66/100/133 RAID controller HPT372; Realtek 10/100Mb/s Ethernet controller RTL8100B; 6-channel AC'97 sound; USB 2.0; etc; ATX. $150-$160. For more details check this and this news-stories.
Kurouto Shikou/Albatron PX845PEV Pro: i845PE; 3 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 5 PCI; 2-channel ATA-33/66/100; 3Com 10/100Mb/s Ethernet controller 3C910-A01; 6-channel AC'97 sound; USB 2.0; etc; ATX. According to this Japanese web-site, the product is priced at $90-$100. For more information check this news-stories.
Mercurio F845PE: i845PE; 2 DIMM slots for PC1600/PC2100/PC2700 memory; AGP 4x, 6 PCI; 2-channel ATA-33/66/100; Realtek 10/100Mb/s Ethernet controller RTL8100B; AC'97 sound; USB 2.0; etc; ATX. $110-$120.
Intel D845GEBV2L: i845GE; 2 DIMM slots for PC1600/PC2100/PC2700 memory; Integrated Intel extreme Graphics core; AGP 4x, 6 PCI; 2-channel ATA-33/66/100; 10/100Mb/s Ethernet controller; AC'97 sound; USB 2.0; etc; ATX. $135-$145.
All mainboards based on Intel's 845GE and 945PE chipsets meet the latest Intel's guidelines and can support the Pentium 4 processors with 3.06GHz clock-speed and higher. They also can take advantage of the Hyper-Threading technology, unlike some of predecessors.
Since the official announcement of the core-logic devices from Intel will happen tomorrow, expect mainboards based on these chipsets from different manufacturers to become widely available during the week to come.
--- And with reports today that intel will into 3GHZ in November, Availability for Christmas looks very good to me..... for intel that is.
http://www.theinquirer.net/?article=5740
Semi
I think a Fed rate cut was already factored into the financial markets (the exact date was probably the question), so I don't know that mortgage rates will change much.
Adam
Maybe not, but if there is another rate cut, I doubt that Mortgage rates will go up, eh? Waiting a little longer has small risk IMO.
Semi
All:
A rate cut tomorrow?
Adam
Meeting date: October 7, 2002
Matters to be Considered:
1. Review and determination by the Board of Governors of the rates of discount to be charged by the Federal Reserve Banks.
Hummmmmm Interesting. Ya know I thought I had "Missed The Boat" (Get It? - Read my other posts) on refinancing the house, and waited too long. Now I think that I'll cancel my appointment with the Bank tomorrow, and wait a little more...
Semi
All this gets back to the question, could AMD assemble a device, burn it in, test it, mark, pack and still deliver it to a customer in a weeks time? I would say yes under emergency conditions but certainly not under production conditions.
EP
Thanks Elmer, that's ALL I was trying to say. I'm glad that SOMEONE got the point, because it appers that OTHERS didn't.
Semi
This is all probably moot, since it seems that the author of that article was wrong. The article is dated 10/3, which is week 41, not 40.
Ummmmm, I just looked at my Badge Calender, and at my company, WW01 was from Dec 30, 2001 (Sunday) to Jan 5, 2002 (Saturday), and so WW40 is listed as Sep 29, 2002 to October 5, 2002. That makes Oct 3 in WW40.
That's just my Company of course.....
Semi
It was you who claimed a device made in WW39 and received by a box maker in WW40 must mean something dire...I claim it does not.
--- I "claimed" no such thing. What I said was......
--- That means that the chip went through Wafersort (Die Functionality Testing As a Whole Wafer), Assembly, Final Test (Bin/Speed Testing as a Packaged Part), Burn In, and Post Burn In Test, in less than 1 week (let's not forget shipping time). This should normally take 2-3 weeks. My guess would be that they skipped the burn-in alltogether, If they did that, and held the testing tools open and unused, and hand carried the wafers all the way through, they might have been able to meet the 1 week timeframe. I wouldn't bet my life on the reliability of that part though.
--- All I "claimed", was that the part was rushed out the door, which is not normal procedure. You're the one who read "Something Dire" into it. If people read into a statement what's not there, it's not my fault.
As to the point that it must mean AMD shipped by air, is there any other way these days? So far as I know, nobody ships chips by boat anymore.
--- Good thing you said, "So far as I know", because what that statement proves, is that some people who have been out of the industry for awhile, don't know as much as they think they do.
Semi
Indeed, when I was at Intel the WW marking was applied at the assembly plant, naturally enough. Now that assembly and test are usually in the same place, with finished units drop-shipped directly to box makers, why would anyone find it surprising that a WW39-marked part could be at a box maker by WW40?
--Tim May
Depends, WW39 is 7 days long. Where these parts made at the beginning of WW39, the middle, or the end? I can't tell, can you? It also depends on where the Boxmaker is. Most parts that I know of are not "Drop Shipped", to arrive at a boxmaker less than 1 week later, as that is the most expensive way. Unless of course you're desperate.
Question, Where do you get the information that "assembly and test are usually in the same place"? Maybe it is, maybe it isn't. Is that the way AMD does it? Certainly Wafer Sort is now attached to the Fab, so that Poor Bin information can be quickly fed back to the Fab, but most Assembly that I know of, is done in Malaysia or the Philippines. I haven't heard of too many Wafer Fabs in those locations. Do you have contrary information? I guess what you're saying is possible (shrug), but it's not how I've seen it done....... Recently.
Semi
Overclockers.com Take on AMD's 2400+ Issues
http://www.overclockers.com/tips00135/
My Take:
Perhaps more importantly, look at the week on the chip. This chip was made week 39. That means the last week of September. We are now in week 40. These chips were rushed, and I mean rushed out. I don't recall there ever being so little gap between week of manufacture and week of sale.
--- That means that the chip went through Wafersort (Die Functionality Testing As a Whole Wafer), Assembly, Final Test (Bin/Speed Testing as a Packaged Part), Burn In, and Post Burn In Test, in less than 1 week (let's not forget shipping time). This should normally take 2-3 weeks. My guess would be that they skipped the burn-in alltogether, If they did that, and held the testing tools open and unused, and hand carried the wafers all the way through, they might have been able to meet the 1 week timeframe. I wouldn't bet my life on the reliability of that part though.
Where Are The Rest?
Look at the facts:
AMD said the last Palomino start was in late June.
--During the summer, the 2200+ was hardly available, only recently have a lot of places been selling them.
--Due to the delays, AMD has acknowledged serious problems with the TBreds that required redesign and structural modifications.
--Very few resellers have offered to even pre-sell either the 2400+ or 2600+.
--They're going to OEMs? Then where are the OEM systems for sale? Neither Gateway, Compaq nor HP are currently selling such systems.
--Per the "new" vaporous Athlons, look here and you have an AMD salesperson telling an OEM making high-end CAD machines that the XP2800+ is going to have a "very limited production run" which has been sold out, so go away and we'll tell you when Barton is ready.
---Shades of intel's CuMine Fiasco?? It's pretty clear that Elmer, and the others that were saying that AMD had sacrificed their 0.13u gate width at 0.18u were true after all. The lack of availability of parts from a Vendor with < 20% market share COULD be explained away, IF..... a person had their head in the sand, but otherwise, the data seems to speak for itself. If AMD could make them, they'de be available.
This is a company that can tell customers, especially the kind of high-end customers they're trying to get through their AMDMe ad campaign "Go away, we don't need your money?"
Does that make ANY sense to you?
--- Not much to say here.........
I could easily be wrong on the whys, but something seems to be going terribly wrong in Dresden based on what's not coming out of there.
This piece may look very foolish two or three weeks from now. Personally, I hope it does.
But what if it doesn't two, three, four weeks from now?
Ed
---No Ed, You're not going to look foolish. Hope is a great thing, I hope I win the lottery, and marry Michelle Kwan (Love those skater outfits), but I doubt those things will happen........ Either.
Semi
AMD Introduces New Process Technology - SOP
---Courtesy of PE, by way of Elmer... On Another Thread.....
I also hear that AMD is working on a new super secret chip, code named...... Origami
Wednesday October 2 , 11:59 pm ET
AMD engineers have been working long, hard hours in the laboratory to improve AMD's technological capabilities and keep pace with their competitors as well as their own marketing requirements.
To that end, AMD has developed a new wafer technology that will be used to produce the new products introduced by AMD over the past few months.
This new technology is a variant of the Silicon on Insulator (SOI) approach. The new material that AMD has harnessed for the insulating layer is PAPER - and the new process is appropriately dubbed SOP - for Silicon On Paper.
"Silicon On Paper is the perfect technology for producing our new paper products" says William Siegel, AMD's Chief Technology Officer. The performance of AMD's Athlon processors on SOP technology is highly scaleable, according to Siegel, and by proper adjustment of the Paper Properties, speeds of the CPUs built on this processor can be tailored to whatever AMD marketing personnel need to compete with Intel.
A new wrinkle to this technology, one that adds flexibility to the design engineers using it, is the full erasability of the programming elements used in the device logic structures. In fact, each processor comes equipped with its own eraser, similar to that of a number 3 1/2 hard rubber eraser, compatible with the paper technology.
" With the new rubber erasability and our Paper Substrate, , speeds and power consumption and even benchmarks can be adjusted in the field by AMD marketing representatives", according to Siegel. The new erase technology allows for no smudge residue, another benefit, leaving no traces of document alteration in case of future potential litigation.
AMD will host a Webcast today to launch the new SOP technology and new SOP processors. According to Richard Heye, vice president of platform engineering and infrastructure at AMD, availability of these new products is immediate and new products can be downloaded from AMD via standard FAX (plain paper is recommended) at 1-408- 982-6164.
Note to Editors: AMD, the AMD Arrow logo, and combinations thereof are trademarks of Advanced Micro Devices, Inc. in the United States and other jurisdictions.
--------------------------------------------------------------------------------
Contact:
AMD
Morris Denton, 512/602-2685 (Public Relations)
morris.denton@amd.com
Mike Haase, 408/749-3124 (Investor Relations)
mike.haase@amd.com
Intel Opens First European Microprocessor Research Lab
Result of collaboration with Barcelona's Universitat Politècnica de Catalunya
BARCELONA, Spain, September 19, 2002 -- Intel Corporation today opened its first microprocessor research and development center in Europe, Intel Labs Barcelona (ILB) located at the Universitat Politècnica de Catalunya (UPC) in Barcelona, Spain. This arrangement will help expand Intel's industry leading capabilities in microprocessor research and development, as well as enable Intel to continue delivering industry-leading performance on future generations of microprocessor.
http://www.intel.com/pressroom/archive/releases/20020915tech.htm
--- I think this is part of intel's continuing push into overseas markets. With the US economy stagnating, and PC sales flat-to-down in the US, I think it's important to go where the future markets are. Get your foot in the door now, and it will be good for intel in the future, IMO.
Semi
Good Article On Hyperthreading...
http://www.extremetech.com/article2/0,3973,570431,00.asp?kc=ZDTH10209TTX1K0000521
Semi
DuPont Photomasks To Serve As Infineon Tech Supplier
ROUND ROCK, Texas -(Dow Jones)- DuPont Photomasks Inc. (DPMI) formed agreements with Infineon Technologies (NYSE: IFX - News) AG (IFX) as the German semiconductor systems company transitions out of photomask manufacturing.
That's what I thought.... Hummmmm, With the additional layers AMD added to the back end (With additional Masks needed), could issues with the Masks and/or delivery/development of the needed aditional Masks, be part of the problem with AMD's 0.13u process??? I know it says the new facility won't be ready until "late 2003", but could Dupont already be supplying (or not supplying) AMD from it's own facilities? What does the agreement to supply AMD with masks, say about when Dupont will start supplying AMD???
Semi
DuPont Photomasks Warns of Lower Revenues, Increased Losses
Online staff -- Electronic News, 10/4/2002
DuPont Photomasks Inc. today warned its fiscal Q1 revenue would be lower than previously expected, with an 11 percent sequential drop.
http://www.e-insite.net/semiconductor/index.asp?layout=article&articleid=CA250334&spacedesc=...
--- Wasn't Dupont Photomask the company partnering with AMD for future Masks???
Semi