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.
"What's Inside?"
http://www.theinquirer.net/?article=10930
Why would they want to merge it into FASL if they didn't have something big? Is there something else they can offer?
CJ - One can hope! Change of topic - what do you think of Coatue's chances? See post below:
Paul
http://www.investorshub.com/boards/read_msg.asp?message_id=1300762
wbmw(CJ, chime in!), as I said, no one knows exactly when they'll get to market, but just look at Coatue's acquisition by AMD. If this technology pans out, it has the power to cause an earthquake in the memory and storage industries. Cheap, non-volatile, fast, easy to manufacture. Just think about the instant-on possibilities, even for complete and complex Windows installations. Hard disks would only be used for rarely used, very large files.
Now, as always, the technology has to be translated into a viable product, but the progress they've made is encouraging. Then there's MRAM and various others, but somehow the polymers seem most promising.
Paul
wbmw, there are some radical things happening in memory, and although it's impossible to say exactly when they'll pan out and make it to market, they certainly could in 2 years. If they do, most computers will have way more than 2 GB of memory.
Paul
EP, that's interesting. When AMD did that (new design on new process) folks were down on them about it, apparently with more than a little justification. So what do you think has possessed Intel to do this? Do you figure they're feeling enough pressure from K8 to do it? Last time they got that pressured, things didn't go too well for them, not that it's a given that such a situation will repeat itself(of course, they were at the end of a design, not the beginning like now). But it does make one wonder if there may be a cause for concern...
And if the rumor is true that AMD has worked out the kinks and will start scaling now, the pressure is only going to increase.
Interesting times!
Interesting Opteron block diagam from x86-secrets comparison article.
;p Guess it's all in the eye of the beholders, Eva and Sheena....
That's very likely true, EP, but being that it's on 90nm, a new process, you never know what kind of bugaboos may come up. It is possible that leakage will be an insurmountable problem, after all, I doubt they planned Prescott to eat 100W, so it is possible that something may be involved that can't be solved. Granted, if anyone can do it on bulk, it'd be Intel, but this is pioneering stuff, so who knows....
Paul
CJ, It's probably more related to Win64AMD than anything else. A Linux-based server is one thing, but a Linux-based desktop/workstation is another - they probably are looking for a larger market than that would provide.
Paul
Well, if 2GHz is to be the bottom and they can continue to scale, it's certainly cause for at least minor celebration at AMD!
Golfbum, seems like I remember reading about MOESI(Modified Owner Exclusive Shared Invalid) protocol way back when, before Opteron was released.
Ah, here it is:
http://www.amd.com/us-en/Glossary/0,,238_242~2262,00.html
Can't own up to it, eh, chipguy? AMD quoted the SPEC # for a 2.0GHz Opteron, which is what Joe is talking about.
Paul
spokeshave, re: Dare I say that SOI is starting to reap benefits?
Nah, that wouldn't fit the FUD! ;P
Actually, I'm hoping you're right, although it would be nice if SOI helped the speed at the same time. But I guess that's too much to hope for. But, it could be that for Athlon64, they'll sacrifice some of the power savings(less important in a desktop than in racks and clusters)and tune the process for increased speed. Do you think everyone could live with maybe 5W more total to allow greater speed? Wonder how much the increased power envelope would affect frequency headroom?
Maybe that's the whole reason for this bizarre mess of rumors we've had - still working out kinks in tuning for speed (and enabling unreg DDR400)?
Paul
IBM'x US eServer 325 page up.
http://www.pc.ibm.com/us/eserver/opteron/325/view_models.html
Yes, and some(all?) of the 3.06 P4 benchmarks were w/ PC3200 and 800FSB.
Haddock, thanks for the fix. I forgot to check it after posting.
Paul
146/246/846 benchmark white paper pdf at amd.com...
...with latency info. Can't wait to see PC3200 results!
Thanks to pgerassi at SI
Paul
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/39497A_HPC_WhitePaper_2xCli.....
max69 - ?! Perhaps you ought to try reading again.
(somewhat OT)tek prophet & EP, I don't think it's even that(at least as far as the "Avarage Joe" goes - don't even want to go into the politicians!!) so much as I think most people don't have even a basic understanding of what business is all about.
Now, I don't have an MBA or anything, but I know that an employer doesn't owe me anything other than what they've promised, and even then, if there is a radical change for the worse in the business situation, that's up in the air. People don't seem to grasp that businesses are in business to make a profit, not to provide jobs or to be a tax scapegoat. They also don't grasp that, since the goal is to make a profit, businesses ultimately don't pay taxes, their customers do through the prices charged.
I'm astonished that people seem to believe that the company is somehow responsible to keep providing them a job, even when it's been bleeding red ink for months/years. We have a textile company here in NC(Pillowtex), that has been in business for 100 years and just finally closed its doors, eliminating thousands of jobs in the Charlotte area. One man interviewed said something to the effect of,"They didn't give us a thank you, a howdy-do or a kiss my b*tt. I think we need to get together and tell them they can't do this!" Now, excuse me, but this company emerged from Chapter 11 bankruptcy a year or so ago, and has never been able to make a profit since.
The textile industry in particular has been "shipping jobs overseas" for at least 10 years now. Frankly, I have little problem with that. It is a low-skill jobs type of business that cannot hope to compete with companies overseas that can find people who are thrilled to do the same job for a lot less money, and often do it better as well! People have known for years that textiles are not going to continue in this country, but they refuse to take responsibility for their lives and learn a new trade. Then they want to blame the company, and the executives, whose job it is to make sure the company makes a profit. Now, don't get me wrong, I don't for a second think that executives are clean as the wind-driven snow, and I do think they/the company have a responsibility to carry on their business in a moral way, but they don't have the responsibility to stay in business, bleeding money, as long as possible, just because their employees want them to.
Now unions are another whole issue, but I won't even go there other than to say they once served a legitimate purpose, but now are corrupt and vastly overpowered, both in their power over their people, as well as politically.
Sorry for the rant. And in case anyone wants to try to compare AMD to such companies, don't even try - theirs is a legitimate enterprise, with a great opportunity to make a great profit, although they haven't lately. The money is there to be made, though, and there aren't too many people overseas who can do what our EEs, etc., here in the semiconductor industry do, much less do it better or cheaper.
Paul
EP, do you know if the P4 Xeons w/ L3 all have the same size L2, and if so, what size is it?
Paul
EP, I didn't know that, but I do assume they meant L2 because of the group it's in. Odd.
drjohn, better look again. The $690 3.06 Xeon has 1M L2 cache w/ no L3 cache. The fastest listed with L3 cache is the 2.8GHz. You looked in the wrong group. Look at the one right above it, and you'll see.
Paul
drjohn & neye_eve, re: tecchannel test
There were both uni- and dual-processor tests done.
The price list they used was screwed up. They quoted the 3.06GHz Xeon w/ 1M L3 @ $690, whereas the Intel price list they quoted says the 3.06 w/o L3 is $690, and there is no 3.06 w/ L3 yet. The 2.8GHz w/ 1M L3 is quoted at $3692!! Insight's prices are even higher(don't have the 2.8 w/ 1M L3yet, either). Hmmm... interesting price/performance comparison.
http://www.tecchannel.de/hardware/1164/3.html
http://www.intel.com/intel/finance/pricelist/
http://www.insight.com/web/apps/nbs/index.php?C=C-Processors&IS=on&M=IT&SB=phl
chipguy, Any company that needs its main competitor
to stumble to make any money at all doesn't seem like a good
investment to me.
So, you spend time here... uuh... why?
EP et al, AMD's only choice is to battle Intel on all fronts. Otherwise, Intel can merely cut prices in whatever market AMD competes in, and use their market dominance in the other segments to keep prices high, thereby subsidizing the competitive markets. AMD is not going to start making video cards, they have expertise in a couple of places, and that's where they are going to compete. Sure, they can try to focus on a particular CPU market or on flash, but then Intel would just choke them off while they continue to make profits in other segments. How on earth can you figure that is better for shareholders than taking the difficult but, ultimately, more profitable path? They may fail, but the failure will certainly be a lot quicker if they do what you're suggesting. Intel is like a shark - it's a fierce competitor, nothing personal, but it's gonna eat you just the same if you give it the opportunity. You have to defend on all fronts, or the shark will just attack from a different direction.
Paul
I don't know if everyone is aware of this, but in NC(my state) and GA, this weekend is a tax-free holiday. I don't know what other states have similar situations. Computer systems, peripherals, and software are part of the exempt list. Could make for a big boom in the back-to-school surge. There are some "restrictions" - such as $3500 for a computer system!
Paul
Joe, re: capacity
The important thing for Intel would be to put a cap on AMD's capacity, meaning Intel would not have to be waging price war for market share, AMD's limited capacity (in case AMD was forced to follow Intel and offer 2 or 4 MB cache on their chips) would do it automatically.
Do you think perhaps that's why Jerry said the outsourcing restrictions were not a problem anymore? Perhaps Intel, knowing their plans for chip expansion/fab buildup as well as AMD's limited capacity, just decided to nix the restriction and to accomplish it another way? Of course, if AMD can pull off Opteron and ramp up the speed adequately, it'll be an expensive "backfire" for Intel!
Paul
eServer 325 TPC results. Wow!
Thanks to Jhalada's post on SI.
Paul
http://www.tpc.org/tpch/results/tpch_perf_results.asp?resulttype=all&version=2%
PC Watch coverage of IBM Server 325.
http://pc.watch.impress.co.jp/docs/2003/0730/ibm.htm
Paul
More benchmarks. Would be nice to see some results from the new Xeons or Itanium there.
http://www.theinquirer.net/?article=10754
http://asdp.bnl.gov/asda/LSD/Benchmarks/benchmark_solve_2.html
Paul
IBM Server 325 Opteron Linux cluster benchmarks. DB2 news, too.
Could someone knowledgable in the area explain the significance of these and place them in some context?
http://www.theinquirer.net/?article=10752
Paul
Opteron scaling is quite nice here:
http://www.amd.com/us-en/Processors/TechnicalResources/0,,30_182_863_8800~72730,00.html
Paul
wbmw et al, is there any reason to think that the "heavy lifting" in creating drivers has anything to do with their "64-bitnes", or could it not be just as easy to port them as developers have said it is to port their apps? Besides, the types of equipment which actually require a specific driver rather than what MS will provide, are the types of things that will want to have both the performance and the bragging rights that will come from the Opteron/A64 systems. I don't see video card makers and maybe storage or networking controller makers hesitating to take advantage(assuming they can point to a performance advantage), they know about marketing, too!
Paul
EP, large cache CPUs?
The label's what matters. Some were made during transition from one package type to another, or it's possible they may be packaging them in two different places, using the two different package types.
paul
Joe, I figured you were probably being conservative. Without specifics to deal with, I figure 5-10% is probably about right.
Paul
Joe, the 3.4% estimate would be quite low, since the change would be from DDR333/ECC to DDR400/non-ECC, not from DDR333/non-ECC to DDR400/ECC.
Windsock, re: your SI criticism of AMD re: Sanders chair at ECE / U of Illinois ...
You don't think it might have something to do with the fact that Dirk Meyer is a alumnus of U of Illinois, do you? And that the school puts out excellent engineers? No, it must just be that darned desire of theirs to throw away money.... not.
Paul