Slashdot Mirror
Intel's Answer to AMD's Hammer - Yamhill
bdolan writes: "Today's San Jose Mercury News is reporting that Intel is going to put a 64 bit architecture extension in upcoming Pentiums if it turns out the Itanium doesn't take off. Hmm. Apparently they intend to only turn this on if AMD's 64 bit processor make major inroads against the Itanium architecture. Aren't we glad that competition is keeping everyone on their toes."
AMD Guy: Hehe..check out my incredible new processor. It's called the Hammer! What do you have in your box?
Intel Guy: Oh..er..I have a *unintelligible*
AMD Guy: What is that? Mumblican? Speak up!
Intel Guy: *coughYamhill*
AMD Guy: YAMHILL? Buwhahahahaha! Intel marketing loves you!
Intel Guy: *cry*
And to think, even as recently as a year or two ago, Intel was being called a monopoly by the FTC and anti-capitalist socialist greens.
If this isn't proof that all "big businesses" can be affected by smaller ones, and to let consumers make and break businesses, rather than regulations, I don't know what is...
Innovation IS CRITICAL to progress. Consumers also want a good product at a price they can afford. While I personally haven't had much luck with AMD products, I know a lot of people who have, and I commend AMD on doing something by themselves that many socialist (democrat) Americans wanted the government to do -- make Intel realize they're not the only fish in the sea.
When they say that they will make this but hope never to turn it on, I can't believe they mean they will put it into the chips but disable it, but that's what it sounds like.
Presumably they mean that they would have the design ready to add to the chips very quickly should it prove commercially necessary.
It's nice to hear they have a backup plan. I've always liked intel chips better than AMD for some reason. (Yes I know I'm probably the only one, and I know there isn't any good reason to so don't flame me for that).
Sig is taking a break!
This has been the focus of some stories at the Inquirer as well.
Personally, I thought that Intel would have been in a good position to just relabel the Alpha 21364 as IA64 and be done with it.
"Provided by the management for your protection."
I wonder if Intel is seeing what AMD saw over a year ago. Many people are looking at the latest greatest operating system and going... oh. That's nice. Does it run my old program? It doesn't? How do I get my Win98 back on there so it will?
Non-backwards compatibility was supposed to be a *benefit* for their new chip.
And now they're suddenly looking at backwards compatibility? Give it ten years *after* and they'll probably be able to *use* a non-backwards compatible chip.
Score one for AMD's clear thinking. No wonder they're breathing down Intel's neck.
No! It's a *SIG*. Keep the Special Interest Groups away! (Con joke!)
what's with intel's names? celery.. err.. celeron, now YAMhill... where's the beef?
Pentium.
Itanium.
Thunderbird.
Windows $YEAR.
Duron.
Celeron.
Boso..err, wrong field.
Okay guys, I don't know about you, but, holding with my "ooh, blinkenlights" philosophy, I miss the days when you could properly identify your processor as an [80]486DX266, and not be overtly pedantic.
I mean, we've even taken a step further in the wrong direction - now AMD doesn't even specify processor Mhz! *WAH!*
Haven't they heard of pipelining and superscalar architecture? Is that statement a result of:
- Intel's marketing folks having no clue
- SJMN reporter not doing his homework
It's quite possible that this processor family makes more advanced use of superscalar architecture and multiple pipelines, but statements like his portray a false idea. I bet we won't see a retraction."What is the sound of one belly slapping?"
I always wondered why they didn't just put three or four processors on a single chip and have instance multiprocessing. I'm sure they would be able to share some of the components that way and reduce the transistor count below what several separate cpus would costs.
And interprocessor communication and cache coherency control would all be on the same chip and so probably easier than normal multiprocessor design.
There is probably a good reason I don't know about so it's a good thing I don't design cpus for a living.
Sig is taking a break!
put a 64 bit architecture extension in upcoming Pentiums if it turns out the Itanium doesn't take off.
You know, the more I've heard about Intel's exciting new architecture over the last few years, the more I think someone's been embezzling the R and D funds, and they don't have a goddamned thing to show for it.
"Johnson, did you finish designing that processor yet?"
"Johnson's not here, sir. He's on a research trip to Barbados with Jan from marketing."
--saint
Wasn't Yamhill one of the hobbit names in The Shire?!
If it doesn't take off? It takes years to develop that kind of new architecture. By then AMD will have it swept.
Don't follow AMD. X86-64 is a follow on architecture, and whatever Intel comes up with wouldn't be much better even if it was different. Computers need to move away from that old decrepid IA32 instruction set eventually.
Intel has a new road and it is not entirely stupid. They are facing the same problem that everyone trying to compete with them has been facing for a long time: compatibility with the installed software base. Either you're compatible and can run IA32 or you're not and you have to come up with lots of other software (enter open source).
Eventually, CPUs needs to move to better architecture. backwards compatability is good during transition, but shouldn't hold you back too much. Go forth Intel and do what everyone else has had to do for a long time, (gasp) struggle for market share.
There is nothing so silly as other peoples traditions, and nothing so sacred as our own.
Well I have developed on ITANIUM, (IA64) It leaves some to be desired, it is a first gen 64 for intel in the consumer market though. I ported BOCHS to the
Itanium, the result can be seen here This may sound loopy at firt but when you look at the backward IA32 incompatibilities, I need a way to test those from within the SAME enviromet.
The IA64 is a pretty lame first attempt from Intel, In my opion, I actually unlike others who will comment have direct experience, I should be getting access to a Hammer shortly, I have heard VERY good things, AMD's effort is much more likley to be a success for several reasons,
But the point I am trying to make is it looks like intel has really dragged its feet here, it cant decide if this is a market to be in or not, If AMD come through as I expect they will Intel will have a HELL of a time playing catchup.
AMD will play to a MUCH broader market than intel can envision, YES I WANT ONE ON MY DESKTOP, And Intel dosent see that market exists YET, then again Intel has never pushed bit copmputing capability, it has almost always lagged at LEAST 2 generations (16 bit when 32 and 64 were availabe) Some of this is vendor support, some of it lack of commitment to it, look at the clock speeds on the Itanium's and tell me, do they really expect this 64 bit pig to fly ?
Sig went tro...aahemmm.....fishing........
Does anyone know how this new architecture would compare with Itanium? I know AMD doesn't really have a dedicated 64-bit architecture. I'd appreciate it if someone could provide some info.
For those who care, Yamhill is a small town WSW of Portland (the little red star at the lower left).
Fascinating info can be found at cityofyamhill.com, naturally.
I don't believe M$ is a monopoly. The only monopolies we've had historically are ones where the government either mandated a private corporation (telcom, energy, etc), or the government subsidized one corporation and tariffed, penalized, or regulated its competition (Standard Oil, etc).
Microsoft has many MANY MANY competitors -- the varieties of Unix, the Apple O/S's, etc. The fact of the matter is, the market and the businesses and the consumers PREFER Microsoft's products. I've tried for years to find a product that runs better, faster, and is easier to use than Office, and I have yet to find one. Netscape over IE? Netscape was a P.O.S., on ANY OS I ran it under.
If your competitors make crappy products, its their own fault. Eventually, M$ WILL HAVE THEIR DAY. They will get hurt, just like Chrysler did without Government intervention, just like many others. Look at MS Network, what a (billion dollar) failure that was.
OTOH M$ keeps the Computer Consulting industry in business. If everything ran well, do you think the industry many of you is in would be as healthy? Thank God for Nimda I say! Job security for geeks.
I am laughing at there choice of the yamhill river for the naming. I live about 3/4 of a mile from the river. They find two headed fish in it, I don't even want to know what will be found in the intel processor.
Are you saying Athlon XPs are not backwards fully X86 compatable!?
That's a rather extrodinary claim, and one I'd never heard anything about before. Do you have any sources you could refrence? The only thing google turns up is info on Athlon XP mobos with backwards compatable PCI slots that work with non-ECC DRAM.
Or are you trying to say Intel's chips are not backwards compatable? I find that equaly unbeliaveable
autopr0n is like, down and stuff.
Why do people insist on wasting their moderation points on "funny" comments?!
The chip is code named Prescott. From the article:
The Yamhill features are being built into the next version of Intel's Pentium chip, code-named Prescott, with an option to turn the features on or off. In 2003 or 2004, when the Prescott chip is expected to be available, Intel will evaluate AMD's offerings and the success of the Itanium and then decide whether to activate the Yamhill code.
There you have it.
I'm a 2000 man.
Well it depends on how you look at the definiation of 128 processing.....
If you mean that the mac can handle a number that is 128 bits in length. Then no, they can only handle 64 bits. Most of the time they only use 32 bits of data though.
However, if you want to know the maximum amount of DATA that a G4 processor can handle at a time, then yeah, 128 bits is correct. Because the G4 has the 'amazing' altivec unit that can process 128bits of data at one time. The lengths of the data can vary (8,16 or 32 bits), to fill the 128bits of processing.
or just the Pentium version, 63.99999999 bits?
John 17:20
I've tried for years to find a product that runs better, faster, and is easier to use than Office
Try Cetus Wordpad.
It's been a long time.
> ones where the government either mandated a private corporation
.. in fact, it's pretty much the other way around. Companies are successfully changing the laws in our countries, with very little public knowedge.
You display your ignorance here. You're not honouring the reality that since we, the people, have been more than happy to chip away at our goverments' ability and legal powers to mandate, regulate and punish (an idea that seems to make most rabid free-markerers piss thier pants in fear). Even a passing knowledge of the changes in trade laws and treaties over the past 40 years would allow you to comprehend that companies have more legal rights and powers on the international market scenes than governments themselves. It's real. People don't want to believe it, but it's real. Read up on NAFTA. Read up on any of the recent lawsuits being launched against governments world wide by private corperations, both domestic and abroad. The point is, it's harder than ever for a government to actually regulate the market or a company, due to the enormous size of corperations (and thus their economic leverage), and their successful con of the public at large in convincing Joe Blow that the government is a corrupt, antiquated insitution that does nothing but collects taxes and wastes money. In short, there is neither public support nor legal support for governments to control the markets much, even if they wanted to. The MS case is a good example of this. Another good example is of a Canadian company suing Santa Monica for 1.3 billion dollars in punative damanges, because Santa Monica was forced to buy 80 of their drinking water at a cost of 3 million dollars per year becuase this company's unsafe product contaminated dozens of free water wells. The State of California (along with 9 other states) has banned their product, and thus, is being sued for it. See? It's way beyond governments regulating anything right now
"Old man yells at systemd"
I don't believe M$ is a monopoly.
Legally, they are. Common sense also says that they are a monolpoly.
The only monopolies we've had historically are ones where the government either mandated a private corporation (telcom, energy, etc), or the government subsidized one corporation and tariffed, penalized, or regulated its competition (Standard Oil, etc).
Huh? Pray tell, where was the Government Mandate or Government Subsidy in the United Shoe Machinery case (to pick one past monopoly)?
United Shoe Machinery (USM) had between 75% and 85% of the shoe machinery market. USM refused to sell it's machinery but only leased, on ten year leases. It also compelled leasees to agree that if they required an additional machines they must lease from USM. USM also provided free maintenance to their machines (or, alternatively, the lease cost included maintenance). The court found that the restictive lease and the free maintenance were barriers to entry by other companies, and removed them from the agreements.
Not a hint of mandate or subsidy here, yet USM were clearly a monopoly (which is quite legal), and were using that monopoly position to quench competition (which is quite illegal).
IBM's power4 chip has 4 processing cores on a chip. Intel and Sun have plans in the works. Intel will do this to follow up with the IA-32 Xeon processor. Here is a story on this
Yes but every time I try to see it your way, I get a headache.
AMD got ahead of Intel on 64-bit with backward compatibility to IA32.
So when Intel releases Prescott and turns on the Yamhill features, AMD's 64-bit system will suddenly be incompatible with Intel's 64-bit system.
There is no chicken and egg, here. Intel will still sell more chips than AMD regardless of compatibility design; then those interested in compatibility will choose Intel to get the larger market to sell their SW into. This will also happen if Itanium prevails, though AMD will have the backward compatibility to help it a little with some markets.
Intel will win, no matter how many people say on message boards they want AMD.
The apt comparison is Microsoft and Apple. Enthusiasm and commitment are not the dominant forces of economics.
--Blair
The diffrence between the pentium and the p-pro are rather minute when compared with the diffrence betwee any pentium/486/386/etc chip and the Itanium. To really answer your question, though you kind of have to look at the history of the whole thing.
:P)
To start things off, intel releases the 8086, and the cheaper 8088 (8086 with a 8, rather then 16 bit bus interface). And thus begins the x86 era.
A little later intel decides they need a 32 bit CPU, but rather then design a totaly new chip, they just add a bunch of extensions to the 16 bit one. They call this new chip the 386, and it's supposed to revolutionize everything. The chip is totaly backwards compatable with the old 8086's and 286s (so the old register AX becomes EAX, but you can still access the first half as AX).
for a long time (windows 3.1) most software still ran in 16 bit mode, not really utilizing the software. IIRC It wasn't untill windows95 and NT started getting used that people really started to take the full potential of their machines in every day tasks.
Now, this is also around the time of the Pentium and the Pentium pro. The pentium ran both 32 and 16 bit software quickly, but the ppro ran 32 bit software faster, and 16 bit software more slowly (of course, the p-pro core became the pentium II, then the pentium III and ran at much higher clockspeeds, so it eventualy became a non issue, a 1.3ghz pIII is going to crunch 16 code faster then a pentium233mmx no mater what
Now, when you look at what AMD is doing and I guess intel now with their rather odly named Yamhill is taking the orgional design and adding 64bit extensions the way they added 32 bit extensions to the 286. EAX becomes RAX, and you can get at the first half by calling it EAX and the first quarter by calling AX, etc.
Itanium is a totaly diffrent thing, it's a whole new system with x86 support tacked on extra, rather then tacking on 64 bit support to an aging archetecture.
Hrm, I hope that explains things.
autopr0n is like, down and stuff.
I don't know what does. If you're a consumer who's pro intel and you've been waiting for YEARS for itanium to be released. Now, Only for it to be usurped by a stop gap processor to compete with a rival. My god. It's better than the nothing these guys have been waiting for and it will be ESPECIALLY painfull to those software houses who have been porting their flagship product to Itanium for some time now... If this chip yama..whatever is made it won't be a dog because they HAVE to compete not only with the marketshare in the 64 bit arena but more important is the MINDSHARE. AMD delievers and Intel doesnt and they both run Windows. We could see some real interesting things come out of Intel. It would also confirm the rumors I've been hearing for some time now that Itanium is dead after Mickenly.
Peter
www.alphalinux.org
In Tracy Kidder's classic book The Soul of the new Machine he discusses the creation of a new computer at Data General, to succeed their 16-bit Eclipses to a new 32-bit architecture. It was shockingly reminiscent of this case, Intel's transition from 32 to 64 bit machines.
In the book, Data General starts to design a fabulous new machine, breaking new ground in a lot of areas, when going to 32 bits. This new effort was called 'The Fountainhead Project', and had all of the best and brightest engineers working on it. At the same time, the hero of the book, Tom West, instituted a new project to do a simple extension of the Eclipse architecture, in parallel.
There was massive infighting between the two camps, and West had to fight tooth and nail for every scrap of resources to build the 32-bit Eclipses; to the point that the machine was almost entirely designed and built by kids fresh out of college because that's all he could afford.
Needless to say, the FHP failed, and Data General released West's machine to reasonable success.
The similarities here are almost eerie, except that, of course, Itanium actually made it out the door.
If you haven't read Kidder's book, it's definitely a great one. Beautifully written, and while the technology has changed dramatically over the last
fifteen years, the social and business rules are still the same.
I love Mondays. On a Monday, anything is possible.
How is what you're suggesting different from Hyper-Threading or "Jackson" technology?
What's the point of adding more bits?
The absolute amount of memory which can be addressed in 32 bits, sans tricks, is 4GB. That's combined memory and swap. Quite a few people care about that kind of thing, namely just about anyone who runs any decent sized server.
Further, consider the rate at which system memory has been increasing, and project it a few years. If it continues, and I realize that maybe it won't, there's a problem.
C//
My favorite part was where they said the Yamhill guys are working to ensure compatibility with AMD's 64-bit vision!
Fascism starts when the efficiency of the government becomes more important than the rights of the people.
In a press conference earlier today, an Intel spokeman confirmed rumors that their latest processor, the 64-bit "Yamhill" is manufactured not from traditional silicon, but is made entirely from yams.
In particular, they seem much better than "Pentium" and "Celeron". Years ago people really berated Intel's marketing for using those names, but I guess everyone's so used to hearing them now that they've forgotten how awful they are.
One of my companies maintains about 600 computers within 15 different organizations, all running different, badly written software. We get maybe 1 call a month about a BSOD, and even that's overstating it...
People don't call tech support as much anymore about BSODs. 5-10 years ago, they were a new thing, now everyone knows to just reboot, even the idiot newbie in the mail room.
I've had enough abrasive sigs. Kittens are cute and fuzzy.
I don't understand why anyone else didn't make a machine that could compete with theirs?
You clearly did not read the comment to which you are replying.
No one else _could_ make a competing machine and succeed, because the other company's contract prohibited people from buying a competing machine.
This is (drum roll) against the law!
Actually, this one might surprise people:
The WTO has the legal powers in place to enforce foreign investor state dispute judgements, (read: governements being sued by companies) and do so. A company can get their case heard and settled in under a year.
The UN can judge on human rights violations, but hasn't one single way of attempting to enforce their judgements. There are simply no international treaties in place to ensure the enforcement of human rights violations. They nailed Peru on wrongfully jailing a woman under terrible conditions for 10 years. They told Peru to let her out last year. She's still in jail.
"Old man yells at systemd"
not to start a flame war, but i have been reading your posts and feel a need to speak up. also, i am probably way off topic.
i am all for competition and think people should try to make a better product. but it seems that the only reason you are willing to accept for why a monopoly might exist is that they make better stuff. companies have made products that were as good as or better than windows.
take Be. they made BeOS, which people still use even though it is dead. microsoft crushed it as it was just getting off the ground. they didn't just out-design Be, they told their vendors that they were not allowed to sell computers that did not also contain windows. microsoft also required them to diable BeOS by default. i fail to see how that makes windows a superior product. maybe it didn't have all of the features of windows, but operating systems to take time to develop. if they are stamped out in their infancy all of the innovation they might have had is lost. things like this also serve as a warning to others who would enter microsoft's turf. apple and the smorgasbord of *nixes survive because they are in different markets.
they innovate to the extent that you will be enticed into upgrading. make it cheaper in the short run to win in the long run.
you probably shouldn't have read this.
This is really similiar to what Intel did with the original set of 80486sx's...
Of course we all know the difference between a 486 SX and DX was the fact that one did have an floating point co-processor and one didn't.
What is not bandied around so much... although it should be... is the fact that the SX and DX chips were structurally identical. The FP coprocessor was simply 'turned off' either on purpose or because it was non-functional on the SX chips. Co-processor chips for SX chips were in reality complete 486 DX chips that cirumvented the SX core.
Will this be the same case with the Yamhill? If it is, will there be a simple (Pencil-overclocking?) method for enabling the extra processing units?
I doubt it.
The next Slashdot story will be ready soon, but subscribers can beat the rush and slashdot the links early!
How is RDRAM a benefit without pulling fancy tricks? RDRAM has both crappy latency and narrow bus (16-bit). Sure, it can clock high, but that means nothing when the bus is 4 times as narrow as SDRAM.
XML is like violence. If it doesn't solve the problem, use more.
>there's the Willamette (a major river, incidentally one of only a handful in the world that run south to north)
Literally thousands of rivers, big and small, run south to north in Siberia. In Arctic Canada, probably, too.
Loks like your definition of the world is USA.
How do you separate the principles from the principle point of the religion? Your reply even tried by removing "Christ" from "Christianity", using "Xtianity" instead. But if you remove Christ, what is left? Are you saying that you're qualified to say which pieces are "good" and which are not?
John 17:20
The Supreme Court have been a bunch of "traitorous" slackers for the past 50 years. The Constitution is very clear on issues like copyright (7+7 years), Social Security (not a federal issue, to be left to the people), welfare (not a federal issue, to be left to the people), the drug war (same), foreign intervention (same), etc, etc. So yeah, I do think we should fire all those black-robed slackers...
The great irony here is the following:
When AMD released the specifications of its upcoming 64-bit chips in the summer of 2000, these ``cowboy'' engineers decided that Intel needed to match its rival. They began developing their own 64-bit extensions to the Pentium line, making sure the code was compatible with AMD's design.
This is Intel imitating AMD, the very same company Intel execs have derided as immitators, recognizing the threat of the upcoming AMD Claw and Sledge Hammers. Another post suggests this compatibility is Innovation. What's innovative, as you noted, is selling something with the big feature turned off. How long before the enlightened OCP weasels figure out how to turn it on and spoild Intel's party?
A feeling of having made the same mistake before: Deja Foobar
Comment removed based on user account deletion
Uh... you mean the government is not a corrupt, antiquated institution that does nothing but collect taxes and waste money?
Have the people really been happily chipping away the governments legal powers? Did "Joe Blow" really make that decision, that we should deregulate everything as soon as possible? I thought it was the government, following the suggestions of industrial lobby groups and economic pundits (not exactly "Joe Blow").
In most of the world, at least, it's not "Joe Blow" who decided that. He's typically promised more government intervention, regardless of whether it makes sense or not. Foreign investors are promised deregulation, regardless of whether it makes sense or not. Government does whatever means more short-term benefits and less responsability.
I don't know. Maybe it is different in the US (I'm not an American), but from what I have seen it doesn't seem so different.
The government (of any country) apparently IS a useless institution. So useless, it chips away its own powers to do anything besides collecting taxes and wasting money. As a matter of fact, most of its current job is passing legislation protecting itself from having (or being able) to do anything else. It makes everything easier.
For the record, I happen to be against government regulation. I have seen way too much of that, and it just doesn't work.
I also think that when for any particular industry the term "government" is replaced by "Microsoft", or any "X Company" that effectively regulates the market as a government (with an electorate of shareholders) it's just as bad.
Freedom is the freedom to say 2+2=4, everything else follows...
Everyone remember the "Intel Inside" marketing campaign? Anyone remember "Authentic AMD"? The Intel Inside campaign was based mainly on FUD and Intel's control over the x86 processor. Since the x86 was a "defacto standard" defined by Intel, only Intel could gaurentee that it followed the standard. If you used other people's CPUs, they might work, but they might not. Better safe than sorry, right?
If Intel publically implements the x86-64 architecture, while more-or-less simultaneously dropping the IA64 architecture, it will be diaster. It would be publically admitting, in deed if not in word, that AMD controls the future evolution of the x86, not Intel. The best Intel could hope for would be for AMD to gain an incredible amount of credibility- which translates as sales in the lucrative but conservative buisness markets. Even worse, the current positions of AMD and Intel might even be reversed, with AMD being perceived as the flagship processor company and Intel the clone maker.
Going to 64-bit is rapidly becoming not an option. Many desktop systems are having a gigabyte of memory installed. Even x86 servers often have three gigabytes of ram installed. The server market is even worse off than the desktop market, as all the ram is generally given over to a single application (Exchange, or a database, for example)- and a 32-bit processor simply can not access more than about 2-3 gig of memory in a single application. The big-iron Unix cpus (Sun's SPARC, HP's PA-RISC, IBM's Power-4, etc) all went 64-bit years ago. It's not unusual to see even "moderate" servers of 4-, 8-, and 16- CPUs having tens of gigabytes of RAM already. The only market that still supports 32-bit CPUs is the embedded market- not a market Intel has ever displayed much interest in.
I figure that the x86 has maybe 3 years to go 64-bit across the board, or we'll be facing another 640K like situation. 3 years is two Moore's Law generations- meaning the people with 1G of memory today will be wanting 4G in 3 years, and the people only getting 256M today will be getting 1G. They'll continue to be hurt in the server market, but they won't lose much in the desktop. Unfortunately, to be 64-bit across the board means the high end needs to be 64-bit within about 18 months (allowing for a Moore's Law generation to push the 64-bit CPUs down the price scale).
Hammer is in a position to do that. McKinnley is the succeed or die point for the IA64. To use an analogy, Intel will have run out of runway- either it flies, or it'll hit the trees.
The successors don't matter- if McKinnley doesn't succeed, Hammer will be there to take the sales. If Intel stays in denial and doesn't offer a viable 64-bit path, they'll be in worse shape than simply admitting that they lost.
At that point, the best thing Intel could do is roll out a Hammer of their own, and plan on less than 50% market share.
Brian
I wasn't saying that the public has been responsible for chipping away at governments, but for the most part, they've stood by and let it happen (or just been to disinterested in existance to care).
> I also think that when for any particular industry the term "government" is replaced by "Microsoft", or any "X Company" that effectively regulates the market as a government (with an electorate of shareholders) it's just as bad.
Agreed, and thats what I'm most interested in avoiding (although I am fully prepared to say that certain countries are very near this situation, if not in it.)
However, I have faith in government (I'm in Canada). I have seen it do many stupid things, but also many good things, which is pretty much no worse or better off than even my 'favorite' companies. And if I ever run out of money, at least I still have a way of expressing my confiendece in the current captains of that boat with my vote, as opposed to the private-interest authority who can and will only listen to those members of scociety who are already in a position of personal security.
The government is my firewall. It may slow things down and cause me problems once in awhile, but I value a last line of defence over unfettered technological advancement and trade any day of the week.
"Old man yells at systemd"
With the VIA 266a being the fastest chipset out there.... Also being rock solid stable way beyond the 133(266DDR) with the epox boards going to 200(400DDR) with good ram chips
The most interesting thing I got out of this article was near the end. They mention that the 8086 was Intel's "backup plan" twenty years ago and that it was designed in THREE WEEKS! I think we finally have an explanation for why the instruction set is such a pain to work with.
But, they just want to be the computer industry's lil' shweet pertater.
--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
Do you realize that your 136-node processor would draw 4-6 kilowatts of power (and so have to dissapate the same amount of heat!), depending on what processor architecture was used?
Would you name all the popular programs you can that scale well onto even 2 processors, and then define the word "parallelizable"?
Would you calculate the amount of time (expressed in trillions of years, exponential notation, or however you prefer) it would take to brute force a mainstream 128-bit encryption algorithm on this cluster?
Are you aware that current sound cards use 16 or 24-bit, 2, 4, or 5 channel, 44.1 (not 144) Khz technology? (I'm probably missing lots of combinations myself).
Would you please do a Google search for "Nyquist", and then explain to us exactly why you want "920 KHz" sound output?
Do you understand now why nobody is willing to "give you a chip plant"?
Do you mind if I use your post as an example, the next time someone else with a 4 or 5 digit UID complains that all the more recent Slashdot accounts are driving the quality of discussion downhill?
However, research carried out in the 1970's, at the University of Manchester, England, revealed that you can do on-the-fly parallelization at the procedural level. It's not easy, but it's possible. There hasn't been much research along this line, for a while, to the best of my knowledge, because there hasn't really been a need for it. (Conventional PVM and MPI are used for tasks that need serious parallel design, and other tasks aren't that dependent.)
What you would do is have the OS distributed over the entire cluster, rather than running in it's entirity on all of them. You'd then use MOSIX-style migration on each and every system call made.
The effect would be to turn serial programs into ultra-parallel clustered ones, with no modification of the binaries necessary.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
In many applications, 64-bit numbers (esp. integers) are likely to be the largest you'll need. They can be manipulated even on, say, 4-bit processors if so wanted, but it will be very handy to use native 64-bit processors. Unix time will probably be defined as a 64-bit integer and the y2038 bug will disappear. Of course, you could probably use a 128-bit processor as a kind of dual 64-bit proc machine, but you could more easily implement an ordinaly dual proc ystem. Can anyone name applications that would require larger than 64-bit numbers?
Escher was the first MC and Giger invented the HR department.
Additions like alti-vec, etc Are mearly extensions and not any program I know of yet REQUIRES alti-vec to run
Apple's iDVD software requires the Altivec unit to perform the MPEG-2 encoding.. I'm sure it could be done on a G3 but the time to perform that operation is prohibitive so Apple requires a G4 processor or better..
A reflective surface radiates far less heat than a non-reflective surface - basically speaking, it radiates at the same rate that it absorbs.
As for your other arguments . . . Well, I'm afraid you really don't seem to understand what you're talking about - that simple bit of physics is a convenient example.
himi
My very own DeCSS mirror.
Actualy I belive most processors can switch between modes, so you can use big or little endian. I think thats the way itanium will work
autopr0n is like, down and stuff.
64-bits should be good enough for a while. The issue isn't the size of integers, it is address space. 32-bits is way too small. There are some neat ideas, such as single-level storage, global/network address spaces, that will eat up vast amounts of virtual address space.
Mea navis aericumbens anguillis abundat
There are many other things that you might want to count besides memory cells.
As another former Oregonian, I will add a little more here.
The Intel CPU code-names are not based upon placenames in Oregon.
They are all *names of rivers* in western North America, primarily Oregon and Northern California (where Intel has most of their employees). The fact that some of them are *also* the names of cities, counties, forests, etc is quite beside the point.
Klamath River (in OR/CA)
Deschutes River (in OR)
Yamhill River (in OR)
Mendocino River (in CA)
Coppermine River (in Canada)
Merced River (in CA)
Tillamook River (in OR)
Katmai River (in AK)
Well, that's all I can think of off the top of my head.
are you completely insane to respin silicon its VERY expensive
lets think about this in terms of actual cost
for intel to actually put this in means that it would have to have 64 bit registers, cache lines for 64 bit + lots of other glue logic this adds to the number of gates and so to the COST you dont put it in unless you have to !
they might tell the marketing droids that its a turn off and on feature but really its not
anyway if Transmeta pull their fingers out they can have the first x86-64 "silicon" out the door as they just have to tweak the software layer
regards
john jones
done 1GHz dual MIPS chip used in network boxs
HP did it with their PA-RISC =descendant of MIPS arch
IBM did it for POWER
SUN dabbled with it for SPARC but the memory coherency beat them they gave up
inetl engineers suck at these things beacuse they are limited by the x86 arch
regards
john jones
Now you are making fun of Yamhill. Not only a river, but a city as well, and a major east-west running street in Portland. If you ever come to Portland, check out Yamhill street. Lots of cool stuff, nice place to get drunk.
Would everyone please lay the fuck off already. We're proud of Intel around here and we're proud of our rivers, cities, and streets. I don't make fun of people who live in New York, even if "York" is a pretty stupid sounding word.
Grow up, assholes.
By Intel and MS at least. On Intel chips it's called PAE (Physical Address Extension). Modern Intel chips support 36-bit memory addressing, if the OS and applictions support it. You can actually buy Intel based servers from Dell with 4-8 processors and over 4GB of RAM. Now the solutions is below average, you have to do windowing like was done with EMS back in the day. But, it does allow for larger amounts of RAM on a 32-bit chip. It's a stopgap until 64-bit stuff hits the mainstream (soon hopefully).
It's already in current Intel chips. They have a modes called PAE (Physical Address Extension) which is a 36-bit addressing mode. Windows 2000 advanced server and datacentre then support this through AWE (Address Windowing Extension). What hapens is an application creates a windows of arbitrary size in it's memory space. It can then instruct the processor to point that window at various parts of the memory using the MapUserPhysicalPages call.
This is a limited measure because:
--It only supports 64GB of ram as opposed to the 2PB 64-bit affords.
--Apps have to be specifically written to use it, as does the OS (2k Advanced server and Data Centre are the only two I'm aware of).
--PAges can't be shared between processes
--A physical page can only be mapped to one virtual address within a process.
--There are security limits on the physical pages.
None the less, it is a temporary workaround, though not one you're going to see a public patch for since app level support is needed.
By pushing their already-obsolete itanic platform (shoving would be more appropriate) by killing anything else that is better (alpha, Mips on NT gone, probably a good conspiracy theory, and any processors that I might forget) and saying how perfect the Itanic is for YEARS, and saying that AMD's road is not a good look at the future and [insert any marketting hype] [insert anything meaning AMD won't cut it with this approach] [insert both point #1 and #2 which means the same thing but change a few words], Announcing this processor simply goes against ALL what they fought for in the last 2+ years. This gives a huge tap in the back of AMD, but at the same time, it's scary because if they don't both use the same extensions, it will segment the market even more. SSE2 is nice, even if I do a lot of 3d, buying a P4 over a dual AthlonMP system won't give me much of a boost, primarely because the rest of the Floating poiont engine in the P4 is terrible when not optimized, and because SSE2 optimization, when implemented fully, will be when the P4 will be obsolete.
Where I am getting at is I don't want to see 2 separate x86-64bits extensions or I'll be really pissed.
Anyways, good for AMD, if they've flexed intel into doing such a move, it shows that they did their homework correctly and Intel probably sat on their cashcow until they had to get their act together.
Too bad the press is all "Intel inside" sold-out... they won't remember how intel pushed against that so hard, now they'll think only present and say "look here's a new processor, wow!!" and be amazed at nothing.
Itanium sucks, what were the last specs? 120W per processor? a pound heatsink? god someone put that puppy out of misery.
--- Metamoderating abusive downgraders since my 300th post.
VILW is an old idea. It's been obsoleted by superscalar processors. It turns out to be better to find parallelism at run-time in hardware than to find it at compile time.
The real reason for the Itanium was to have something that had some intellectual property that AMD couldn't clone, allowing Intel to crank up the price and get their margins back up.
As for the AMD 64-bit machine, it's entirely vanilla. It's very x86 like, with the same instruction set, a few more registers (yay!), and of course the registers are longer. It has all the obvious backwards compatibility stuff. It comes up emulating a 32-bit x86 machine, so old OSs will run, but can be put into 64-bit mode. In 64-bit mode, it can simulate multiple virtual 32-bit machines, so you can have a 64-bit OS running both 64-bit and 32-bit processes. (Run 32-bit Windows under 64-bit Linux!)
Wierdly, the x86 instruction set isn't viewed as that bad today. The variable-length instructions aren't that much of a problem to decode any more. Speculative decode takes care of that. One big advantage of RISC architectures was that making all the instructions the same length simplified decode and allowed more look ahead. That's a dead issue. Making the instructions all the same length causes about a 2x code bloat, which is now unnecessary.
The other big RISC advantage was having lots of registers. Register renaming and caches have killed that advantage. Today, a register is just a short name for a recently referenced variable. There are far more registers inside a Pentium Pro and later than the few explicit ones you can mention in x86 code. In fact, one advantage to not having too many registers is that it shortens subroutine calls and context switches. The machines with huge numbers of explicit registers, like SPARC machines, put a lot of effort into saving and restoring them.
The pre-runtime optimizing, done by a compiler/linker combo can be hard, true, but that's now. Within 10 years, it will be devastatingly good. To start THEN with a VLIW machine, like the EPIC architecture, would be too late.
Furthermore, using a JIT in hardware to optimize the code at runtime (which is what you want, since you can decide what to optimize better and also in what way) is the way to go, Transmeta already uses such a setup in the crusoe.
So: an EPIC based proc, combined with a JIT chip would be ok. When the compiler techniques are better (it's a new area, give them a couple of years, optimizing techniques were focussed on optimizing for a pipeline processor), the EPIC instructionset and the VLIW technique will turn out the technique of the future. Give it some time.
Never underestimate the relief of true separation of Religion and State.
I'm talking about heat that is /radiated away/ from a body - this is typically how heat is removed from a CPU or the like, even when there's a fan.
Take a chunk of metal, paint it matt black, heat it to 100C, then sit it in air at 25C for 20 minutes. Do the same with an equal mass and density chunk of the same metal that has a mirrored finish. You'll find that the matt black chunk has cooled down a lot more than the mirrored one. The matt black surface radiates more energy than the mirrored one.
Look in any first-year university physics textbook for a discussion of black body radiation and related stuff . . .
himi
My very own DeCSS mirror.
True, but you already have native 64 bit floating point numbers and 128 bit integers on P-4. Strange that SSE2 doesn't implement a 128 bit float, though (strange, I say, because the bits are their in the SSE2 registers).
C//