Become a fan of Slashdot on Facebook


Forgot your password?

Pentium 4 Delayed 127

An anonymous reader noted that CNet has a story saying how the Pentium 4 will be impossible to get for manufacturers wanting to ship them over the holidays. Apparently the system makers aren't that happy... but considering what Intel was charging for the things, I can't imagine who would buy one.
This discussion has been archived. No new comments can be posted.

Pentium 4 Delayed

Comments Filter:
  • I bet that this was for computer manufactures to get the product by Oct 30 so they can integrate it into there product line. Not to mention the cuts in cost for the Christmas 2k++.

  • by twingo_gtx ( 161891 ) on Friday September 29, 2000 @04:34AM (#744939)
    Also add to that the current price of the
    dual boards, which makes buying a faster CPU
    still cheaper that 2 slower ones.

    This is no longer the case. You can get dual boards for under $150 US which certainly is cheaper than buying the highest rated processer at any give time.

    2-P3 700's + mobo = 400 + 150 = 550
    1-P3 1000 + mobo = 600 + 50 = 650

    I would also expect the dual 700 config to yeild much better performance for the casual user than the 1000. Now if you argument is that its better to just get the 700 and the regular mobo then yeah, that would be alot cheaper, about half the price, but we're talking about getting more perfomance not less.
  • Intel has been a strong proponent of Vaporware for quite some time now.

    Anyone seen a 1GHz P!!! system yet? I havn't but I have seen actual 1GHz AMD boxes.

    And, remember the 1.13 GHz that Intel released and revoked just to beat AMD (who has released a 1.1GHz).
  • hey, i bought 10 120mhz pentiums for $100 to crunch seti

  • intel is playing catch up in a market that has been around for too long anyways. the x96 isa, as I am sure most of you already know, has been on life support for a number of years now. Yet, the family doesn't want to pull the plug just yet--it may start breathing again any time soon. Look at the PPC chip. Sure, motorola is having some problems bumping it up to current standards but it is still one hell of a chip. It compares favourably to even the best of the x86 chips out there and generates less than a 10th of the heat. AMD *WAS* on the right track when they first introduced the Athlon. At the core it is a risc-esque chip with hardware translation. Too bad the risc-like core is not accessable. Imagine a chip that could do both? Now that I have read the specs on the Sledghammer I am no longer holding my breath though. A risc-like core that maintains compatability to the old 16 and 32 bit code will adding 64 bit code - all to be translated back into AMD's core code.

    It's dead. Let ot go with what is left of it's dignity before you pilleage it all.

  • by Molt ( 116343 )
    It could be worse.. this could have been Pentium 95, that's what normally follows V3, right?
  • I wonder how long they'll keep upping the clock rate on the beast that is x86... I have no doubt that they (Intel or AMD) could build a much better chip if they were willing to break compatibility, but I don't see that happening. I personally don't care too much about the cutting edge of x86 performace; it's just a really ugly hack at this point. If I need a fast machine, I'll use the Sun Ultra80 in the multimedia lab... The IA-32 architecture is good for one thing: making really cheap chips. And that's only because the quantities are so huge and the competition is so tight.

  • Too bad AMD is shipping 1GHz chips when Intel is recalling theirs. :-)

    Intel are certianly the kings of the demo, but AMD is slowly moving up through the royal court in actual production and yield.
  • by NoWhere Man ( 68627 ) on Friday September 29, 2000 @05:01AM (#744946) Homepage
    The power consumption of this chip is too high. I don't want to have to by a separate power bar just for my frikken case (especially when video cards are using external power now too).
    And when you consider that you can get 2 chips and create a dual processor system that can run as fast or faster, you have to wonder why people would want to buy it anyways.
    When they broke the 1Ghz barrier I knew a few people who were already enjoying that speed with a couple of dual 500s running GNOME (Granted you don't get the full 1000Mhz experience, but its pretty close).
    I am still waiting to see a Dual Athlon motherboard, strap on a couple of T-birds, and let those pengiuns fly!
  • I know exactly what you're talking about. The FPGA design for the current project I'm working on takes almost a half hour to compile on my PIII 600. One of my coworkers recently suggested I consider switching to Altera FPGAs for future projects. He said they have a system where customers can submit their designs to one of Altera's dedicated supercomputers for compiling really large chips. I have no idea what it costs to do this though.
  • "Another point to make on this issue is that really high-end CPUs get used for one of 2 things in the real world:

    1. Rendering
    2. Games"

    You forgot one:

    3. CEO's
  • Whatever they do, I'd image that companies paying top dollar for engineers to sit in front of slow software will be among the first in line for faster cpus!

    Best of luck with getting an upgrade. I've worked for a few places where you got upgrades when money was available (budgeted), unfortunately, even though Moore's Law has been known and repeated for years, too many companies still don't get it. Execs are usually the first to get the speedy new box, so they can wave their new and improved phallus in front of other execs. It's beyond me how this improves the company bottom line, but I'm sure it makes sense from where they have their heads wedged. With Dilbert-like logic, a few minutes of the engineers time isn't justified by the expense for a new workstation, but, by golly, they need that new design ASAP

    Perhaps the shortage is due to a shortage of aluminum for those massive heatsinks. ;-)

    Chief Frog Inspector
  • You will need 4 CPU's to run Win2kNet+1/2 Millenium was last year version to come:

    1. 8088 for word processing and email clients.

    2. 80486 for the underlying Windowing system.

    3. Pentium IV for all the online shopping.

    4. Pentium VI for the fully skined talking paperclip of your choice. Insert list of babes here.

    Anything less and response will get chunky.

  • Has anyone else noticed how the power consumption, heat production and size of those modern "microprocessors" slowly reaches the level of ENIAC-type of computers?
  • by Anonymous Coward
    I didn't read the article and I have no idea what this is all about, but in true /. fashion, I have an opinion anyway. It is this:

    Intel $ux0Rz, AMD Ru1z

    There. How's that?
  • If the game's got multiple threads or has a seperate server daemon process, the SMP machine will edge out the uni-processor one if the SMP machine's CPUs are anything faster than half the uni-processor's CPU. Same goes for anything out there that has threads or has external processing applications. I don't want a 1GHz machine- I'm perfectly happy with a dual or quad 750 for now.
  • Damn the FX chipset for the Pentium II!

    Does anyone remember that debacle? Intel planned to release both their PII and the LX chipset simulataneously, but (suprise) ran into chipset delays. The PII was released about three months before the LX chipset. So Intel took the aging FX chipset from the Pentium Pro, and slapped it on their Pentium II motherboard (PD440FX). No AGP slot, No DIMMS (uses 72-pin SIMMS instead). A pentium II 233 cost about $650, and the motherboard was about $270. Three months later they were completely obseleted by the LX chipset for the PII. Even worse, to upgrade motherboards from FX to LX, you would have to buy all new RAM and videocards. Sucks hard!

    That is still my main computer, but all my l33t friends laugh because I don't have any AGP mojo.

  • All hail Intel, king of the pre-overclocked processor!

    For all you people out there who don't want to go through the pain of overclocking your processor, Intel will sell you one that is over clocked before you buy it!
    Just set it to the speed on the box and watch your computer burst into flames.

  • ...and call it "Pentium Squared."


  • Bear in mind that AMD hope to have dual cpu duron and athalon mobos out in the next 4 months, and have 4 and 8 way ath boards out by the end of next year.

    What's taken so long? Athlon and friends use the Alpha EV6 bus. There have been multiprocessor Alpha boards for ages!
  • by jafac ( 1449 ) on Friday September 29, 2000 @09:25AM (#744958) Homepage
    an experiment:

    P3 600, 256 megs RAM, NT 4.0 sp6a, Word 2000.

    Click on word doc on desktop (other apps open, Outlook, Netscape-6 windows, Palm Desktop, Task Manager).

    Word launches in 2 seconds, BUT word doc takes 15-seconds to open and render! Close window; Word takes a FULL 35 SECONDS to close, during which time the window is completely unresponsive to ANY commands. Will not move, resize, or minimize.

    MS software is a peice of poo, to be sure, but I really need a 2 GHz P4 NOW!

    Soylent Green is people!
  • I've been saying this for YEARS, but unfortunatley, I lack the intestinal fortitude and drive to actually do it.

    We need a web site, that works like that tracks all product announcements from major industry players, and charts how well they do or do not stay on schedule, - a betting pool could even be done around this.

    To make things precise, statements like "It will ship in the 3rd quarter of 2001" will be interpreted as the LAST SECOND of the last minute of the last business-day of that quarter. Especially if that preceeds a 3-day weekend, where you KNOW the QA dept will be putting in heavy overtime.

    This way, there would be a PUBLIC place where all businesses could establish their reputation, and their mistakes will not be forgotten. In this manner, vapor will be actually discouraged, and there will then be negative incentives to BS product release schedules and roadmaps - and perhaps some reality will be injected into the picture. Schedules should be set by engineers. Not Marketeers.

    Soylent Green is people!
  • Number of good points I did not realize. First, I did not know about Intel was RISC inside as well. Point also about the Altivec instructs. But, one does have to admit that something should be changing and soon. Being that I am a mostly portably user, I have been looking fondly at the PPC's and their next to nill power usage. Sure, we have transmeta coming, but that is just a good idea so far, not something I would buy. There is alot of inefficiencies in the current crop of x86 chips.

    Oh, and I am not some computer archiect grad. I just like to pretend that I is one ;)

  • by KeyShark ( 195825 ) on Friday September 29, 2000 @04:35AM (#744961) Homepage
    A pentium 4. It's like another rocky movie.... Doesn't marketing do anything at Intel.
  • but considering what intel was charging for the things, I can't imagine who would buy one.

    I'm working on a project with a FPGA chip. It takes my 800 MHz machine about 1.5 minutes to compile the chip's design, using the Xilin x Foundation Software [], for a relatively small design without much synthesis! Even a tiny change to just one gate and I've got to wait 1.5 minutes. It was about 5 minutes before I upgraded the CPU to 800 MHz. My chip is a 10k gate (supposed capacity). It's hard to imagine how anybody can compile designs for the really large devices []. I suppose they use more high-level simulation and don't do as much in-circuit download. Whatever they do, I'd image that companies paying top dollar for engineers to sit in front of slow software will be among the first in line for faster cpus!

    There are lots of other high-end applications like this. The low-cost PC is the bulk of the market, without a doubt, but it doesn't take a lot of imagination to see high-end applications that really need more CPU horsepower. If I'm still doing a lot of FPGA work when the chips are actually available (at $1000 to $1500), I might even upgrade my own home machine!

  • With Linux finally having some decent SMP support and Windows already possessing it (at least in the latest versions)

    Mmmmm, BeOS []

  • by Anonymous Coward on Friday September 29, 2000 @04:36AM (#744964)
    We live in interesting times. Personally I don't really bother when P4 is available but on a side note, I see a trend. The big dragons (MS and Intel) are screwing up big. My interpretation is that what is happening now is marketing Deps panicking, realizing that they are not immortal. Realizing that there are actually other guys out there who can shoot from the hip. Once doubt sets in the mistakes come as well and trying to spank the development process to produce more in less time backfires. My 15 secs being infamous.
  • The MX chips were great. Ran software fine. Chipsets integrated well with it too. Too bad they belong to VIA now
  • Unfortunately, while rendering is helped greatly by two processors, actual modelling only uses one in most 3D packages. Kind of a pain...

    Plus, SoftImage (at least) requires a separate lisence for each processor. Yes, that's absurd. (I don't use Softimage, but know someone who does).
  • I started with Cyrix, on MY first PC. It was a 386SuX 16MhZ.

    Since then, I have been using AMD, Cyrix, and Intel processors. Lately, I have two Intel boxes, mainly because I found Dual CPU boards for fairly low prices.

    Now checking out prices, I think my next purchase will definiatly have to be AMD. I just checked the site:

    AMD - ATHLON - 1GhZ - $444
    AMD - THUNDERBIRD - 1GhZ - $442
    INTEL - Pentium III - 1GhZ - $719

    This seems a bit odd to me, why would I want to pay a tariff of about 61% on a chip that has been rated poorly by many reviewers?

    My other option, if I am going to buy at the Top-of-the-Line, is to buy an AMD 1.1 GhZ for about $60 less than Intel's 1GhZ processor.

    The only advantage to buying Intel is Dual-CPU support. Though if I am not running a server, there's not much point in that.

    you are not what you own
  • by henley ( 29988 ) on Friday September 29, 2000 @04:37AM (#744968) Homepage

    Another point to make on this issue is that really high-end CPUs get used for one of 2 things in the real world:

    1. Rendering
    2. Games

    (I'll ignore things like SETI@HOME / here 'cos noone buys a machine for that, right?....Right?)

    SMP is a fine solution for rendering; I can't speak as to whether common packages out there can exploit SMP though.

    However, very few if any games can exploit SMP. I'm ignoring the subclass of multiprocessing which is used in gaming: specialisation of processing to dedicated hardware (aka 3d acceleration). Partly this is because of the platform (most games are written for that non-SMP OS Windows 9x), and partly it's because SMP has such poor penetration to the consumer market (mostly because of the first reason!).

    So, completely ignoring the very valid question of whether SMP is a suitable model for pushing forward the field of general purpose multiprocessing anyway, the answer to why SMP isn't really a good solution in this case (and hence why Intel/AMD et al can still make news, profits and push out megawatts of combined waste heat) is that the overwhelming majority of systems into which their high-end CPUs are placed couldn't make use of any other solution for their intended use.

  • Ok, This is coming from someone on a Unix/Linux (majority) web site? I think you are mistaken. I do work with a lot of hardware and software and have yet to see an aplication that NEEDS to be run on an Intel chip. Maybe you are confused by the adds for something made for MMX enhanced CPU's? Intel, AMD, Winchip and Cyrix all used Intels MMX technology.
  • The last time I went shopping for a CPU, I found that the "knee" (max(bang/buck)) in the price/MHz curve happened at about 85% of the top speed available.
  • I've got a replacement gigabyteboard with a BX chipset in December of '99, it not a '96/97 original BX so I hoped it would support faster PIII's. The bus speed is configurable up to 133 mhz, so I was hoping that I could get away with getting a 1Ghz PIII in my existing motherboard. Strangely the manual only shows processor settings up to PIII 700 as examples so maybe your right about the 700 mhz limit. Damn those mobo manuals, no where does that mention a 700 mhz limit, but then again maybe I'm just stupid. P.s. wheres a good place to find more info on individual chipsets as I never trust the mobo manufacturers to provide unbiased info
  • I have never seen an SMP system where I live at all. Additionally I would think that paying for *two* processors in a computer would be much more expensive than one at least after a few years of waiting.
  • Go SMP.

    Yeah, that's the answer, for a closed-source single threaded application, that is a compiler for a FPGA chip whose silicon-level details are a close gaurded proprietary secret! You'd have to no only reverse-engineer the software but the chips as well, and the EULA won't allow it.

    Even if the code were open source (or available with a restrictive license), and even if there was some straightforward way to make it multithreaded (has anyone even figured out how to truely utilize multiple CPUs in a C compiler, and make -j doesn't count).... even if you had a reasonable chance of hacking on the code to mulitthread it, if the goal is to save time, buying a faster CPU is probably the best move.

  • Not entirely true! Rendering usually runs on one CPU, yes, but the other can offload a lot of other stuff. Take a look at a CPU usage graph on an SMP box playing a 3D game. Under Linux, one CPU can handle the X server, and the other can take care of the game. This is automatic, since the game and the X server generally require more CPU time than any other process.

    SMP does help for a lot of things. I have a dual PII 350, and while a 350MHz CPU is nothing to write home about, their combined power means that I can play any game that's currently available for Linux (I've tested most of Loki's stuff on this box). I could probably play just about any game available for Windows, but I don't own a copy thereof, so it's a moot point.

    My SMP machine is much more responsive than any comparably-equipped single processor machine I've seen. It almost never becomes sluggish; indeed, one CPU frequently becomes saturated, bringing the load to an even 1.0, but the other CPU is still there to respond to input.

    You're right that SMP has less market penetration, mainly due to Win98's (or whatever 2-letter code they're up to now) lack of support for multiple processors. This has kept it out of the low-end market. I think this may be a self-fulfilling prophecy as well: there's no incentive to add SMP support to the Win95 series, since there's a very small base of SMP users, and that base isn't likely to grow until there's support in the OS.

  • How am I going to heat my house this winter?

    Use an Athlon. They are quite adequate to the task.

  • >> -- Everyone knows /.. is as good as /.
    Oh please. Everyone knows that /.. is the previous stuff, where you've all ready been, what you've already seen. /. is current, where it's at, right here right now. Just slip it into any old path, and where it is, there you are.
  • by Inoshiro ( 71693 ) on Friday September 29, 2000 @05:17AM (#744977) Homepage
    A dual chip system would certainly be good for most "desktop users" who run one application at a time. These users tend to also run operating systems which use cooperative multitasking at some level or another. With two processors, you can have one running their copy of Word, and the other running the base OS.

    I know a lot of people probably have an MP3 player and other applications in their system tray. Maybe they're "not on the screen," but they're still competing for resources.
  • I'll ignore things like SETI@HOME / here 'cos noone buys a machine for that, right?....Right?

    Umm, yeah. A little more than a year ago, I only had a Celeron 366.

    When the Athlon came out, I had to get one so I could support AMD "even if nobody else would".

    The Athlon wasn't very stable, and I couldn't go back to a 366 after that, so I upgraded the 366 to a 450 and delegated the Athlon to a life of Windows 98 and games, which it ably handles.

    The 366 started looking really sad just sitting there without a motherboard, case, hard drive, CD-ROM, or floppy, so how could you NOT cobble together some parts and make the thing run BIND, squid, and Apache?

    So here I am with systems that perform within a few percentage poins of each other. I have a Celeron 366@458, a PII 450@510, and an unstable Athlon 500 that would be faster, cheaper, and more reliable if replaced, rather than OC'd. They all run SETI@home, but I SWEAR it wasn't meant like that...

  • it only goes up to 700 OFFICIALY. You still should be able to run 133 * 7.5 for your ghz PIII.

    Check out websites like, etc. They usually have unbiased info (execpt toms, he hates intel)

  • I don't know of people who complain that Word is too slow on a modern machine. If this is the case - RAM is your answer.

    This will make a difference for doing things like

    browsing the web (one process per window) which will improve tremendously with badly behaved flash and java applets.

    playing quakelike games - one processor for the game physics - one for the bots and the networking

    video - one for the video - the other for the sound and the rest of your PC

    cd writing - on an SMP machine you can burn CD's under load even in Windows. Lots of people complain that their computers lock up whilst burning CD's


    This covers a fair number of things - this will make a difference for many 'normal' users, they will just discover that their PC doesn't appear to slow down when working hard.

  • >Doesn't marketing do anything at Intel. No, thats just it, they spent months comming up with that name. And, I bet they got well paid to do it. I must be in the wrong line of work. Everyone whined when AMD announced 'Athlon' and 'Duron'... but at least they came up with new names. -nite
  • If that's their strategy then it's not working. Micron has just announced they will be making AMD machines, AMD has won over a large system builder in the UK and european sales are very strong. Whether it's Bang for the Buck or supply chain woes, Intel has left the door adjar and AMD has been in position to take advantage.

    With the rumored failure of the Itanium-McKinley, AMD looks positioned very well with their Hammer.

    Chief Frog Inspector
  • Why does this not suprise me. I would expect the P4 not to get any real system integration until March, if not April of next year.

    They're just too damned expensive.

    -Julius X
  • Two words: MicroVAX II.

    I actually fire mine up at work sometimes when the room gets a bit of a chill. With headphones on, you can hardly hear the disk...

  • by Anonymous Coward
    More than your mom will let you dig out of her purse, Slashdot-Terminal.
  • Actually, a bus cannot move faster than traffic. And it has a lot of stops! ;-)
  • This Sucks Because 1ghz pentium 3's are way too slow. How am I supposed to get anything done?

    Maybe I'll Just Suffer

  • by romco ( 61131 ) on Friday September 29, 2000 @04:16AM (#744988) Homepage
    How am I going to heat my house this winter?
  • by Mr Z ( 6791 ) on Friday September 29, 2000 @07:19AM (#744989) Homepage Journal

    The IA-64 does not execute IA-32 via software emulation. They do have IA-32 instruction decoders on the die.

    The main issue is that IA-32 vs. IA-64 is modal, and so you can't mix the 64-bit and 32-bit code with a very fine granularity. From what I understand, it the mode-switch was meant to be thrown with about the same granularity as a context switch.

    Sledgehammer, on the other hand, sounds like it's trying to be a straight extension on IA-32, and so would layer over IA-32 much like IA-32 layered over the 80286, which layered over the 8086... This would allow 32-bit and 64-bit code to mingle within an application. (Just look at Windows 9x for an example of a deployed system that operates in this manner, and why Sledgehammer might hit where Itanium misses.)

    And one last thing: Itanium is the collective name for the IA-64 platform, whether it's Merced or McKinley, just as Pentium has become the name for the current set of IA-32 chips. Merced might get cancelled, leaving McKinley as the first Itanium chip to ship. Wouldn't surprise me in the least.

  • I was wondering how many of my fellow nerds are still boycotting Intel after the PSN fiasco? I will refuse to buy their chips unless AMD does something dumber than that.
  • by flatpack ( 212454 ) on Friday September 29, 2000 @04:16AM (#744991)

    Not meaning to come across as flamebait, but it seems to me that the future for people wanting a high end system is better served if they start exploring SMP options rather than the increasingly flaky vapourware that Intel keeps pushing out. Sure, AMD are pushing ahead with some better quality chips, but why pay all that money for a high end chip when you can get two cheaper ones for the same price?

    With Linux finally having some decent SMP support and Windows already possessing it (at least in the latest versions) it makes far more sense IMHO to go down this route if its performace you're looking for. Even with all the latest advances in processor technology, there's still only so much a single processor can do at once.

  • I don't think I'm gonna try to save up my money for the next 2 years to buy a new case, motherboard, and a chip that will cost 4 time more then the new ones just for a speed boost of *2. Heck I bet starting up a computer with a P4 in it sounds like a freaking jet engine because of the fans to cool the thing. How many people are considering buying the first P4 chips anyways?

    - Micro$oft
  • im still trying to upgrade my pII celeron.....

    "sex on tv is bad, you might fall off..."
  • Hasn't this been the case with the high end Pentiums, Pentium Pro's, Pentiums II, and Pentiums III's ? :-)
  • by grahamsz ( 150076 ) on Friday September 29, 2000 @07:28AM (#744995) Homepage Journal
    Traditionally AMD's market has been the entry level pc one, and they've never had such things to worry about.

    Right now thier main worry is keeping production up to speed to keep intel out of the top slot.

    However most current users of SMP are high end server manufacturers, and most enterprises prefer to use the tried and tested intel xeons in their systems.

    It would have been foolish of amd to focus their efforts on smp when their chip didn't have the respect it does now.

    The good thing is that the EV6 Bus lends itself a lot better to multiprocessoring than intel's architecture. We may yet see a 32 cpu AMD system.... sweet :)

    Intel on the other hand just cant do that until they get their itanium chips rolled out (late 98 I seem to recall :). But AMD on the other hand have the sledgehammer ready to strike back at the itanium.

    The future looks bright for amd so long as they escape any major cockups.
  • Nope. In fact, it's already been installed, and your home is now considered a propritary database by CueCat.

    Any attempt to use the bathroom will be met with legal action.

    Godot called. He said he'd be late.
  • Vapourware has been (ab)used by Microsoft and other software companies to stop concurrence...

    Are Hardware manufacturer using the same methods ? Vapour P4 to avoid people buying AMD processors ?
  • In conclusion, and to repeat myself, the Itanium processor executes IA-32 applications in silicon with complete compatibity with current IA-32 processors.

    And that's the #1 reason why the Itanium is a Sick Little Monkey.


  • Quake III Arena can use dual CPU's on Linux, Windows NT, or W2K. (And if you consider the T&L support on a GeForce or Radeon video card to be a processor, then Q3A supports three "processors", as it takes full advantage of T&L hardware.)

    It's a pretty safe assumption that all the games that come out over the next year using the Q3A engine, like "American McGee's Alice" will support dual CPUs as well.

    I've never had the pleasure of trying it myself, but according to John Carmack and reviewers, dual processors doesn't boost the maximum frame rate too much, but it does really help remove the drops in frame rate that one normally gets in highly complex scenes - like when 5 player models are on screen plus a bunch of explosions, curved surfaces, with gibs and rockets flying everywhere.

    I think that Intel's continuing problems getting high-end CPU's out the door will make dual CPU machines ever more attractive for power users who run Linux or W2K. The price/performance comparasion is amazing - two PIII 700's cost $400, but a single PII 933 is at least $460. Dual CPU motherboards are not much more expensive, either.

    Torrey Hoffman (Azog)
  • This was one of the major priorities for Intel's Itanium project. It will still run x86 on die, but it uses a whole new instruction set and architecture as well. This was a chance for everyone from the high-end down to slowly trickle away from x86. Unfortunately, AMD's Sldgehammer just "extends" i386 into 64 bits, almost identically to how Intel extended x86 into i386. AMD could single-handedly cause the 25 year old architecture to remain in place for years. It's another quick-fix, just like i386 was. Sick of x86? Boycott Sledgehammers.
  • by jmv ( 93421 ) on Friday September 29, 2000 @04:22AM (#745001) Homepage
    The problem is that most users only run one (active) application at a time. Because most of these applications are not threaded (ever seen an SMP version of Word?) you still end up using only one CPU. Sure, I'd know how to make full use of even a 4-CPU box, but most people wouldn't use more than one CPU. ...Also add to that the current price of the dual boards, which makes buying a faster CPU still cheaper that 2 slower ones.
  • 1-P3 1000 + mobo = 600 + 50 = 650

    just where are you planning to get said p3-1000? last i heard they were about as easy to find as an FSU fan who wasn't cocky.
  • > RISC doesn't mean fewer instructions it mostly implies the instructions are simpler

    Actually, that's precisely what "RISC" means, which is why it's a bad term. You are correct that the instructions are "simpler" in the sense that operations are *only* allowed on values already in registers, and memory references are confined to simple "load" and "store" instructions. Thus, memory access and computation are seperated, simplifying the internal implementation. The move to a load / store architecture is atleast as important as the focus on having a small, highly orthoganol instruction set. I guess we call it "RISC" as opposed to "load/store architecture" because it's sounds cooler.

    > Also RISC processor are more register based than memory based; in other words operations happen on registers on the CPU and the result go back into registers on the CPU.

    You are being a bit loose with your description. *All* modern processors pull their arguments into registers from memory, then operate, and then move them back out. In the CISC world that sequence could all be represented by a single op code, but internally it still executes those tasks in that order.

    When papers started coming out on RISC, a lot of designers realized that their machines were already load / store architectures *internally*. Whenever a Cray got an opcode to add two different memory addresses, it would internally break that into four seperate micro-ops. Two to load the operands from memory into registers, one to execute the ALU operation, and one to store the result back to memory. So, at the microcode level, the machine actually operated as a load/store architecture, it's just that the ISA (instruction set) (which was inherited from the earlier PDP machines) was conventional CISC style.

    With the rise of decent compilers and the fall in RAM prices (which made increased executable size less of a sin), it made sense to have the instruction set itself be based on load / store, thus moving all that decode logic that used to be in hardware out to the compiler. Bam: less hardware. Lean and mean.

    Unfortunately, the x86 world never did move that logic out into the compiler which means they have massive decode units on their front ends to this day. Incidentally, Intel's IA-64 is a VLIW-style architecture which can be seen as the next step in hardware removal. In a VLIW system (or EPIC, as Intel calls it), instructions are issued in "packages" that are guaranteed to be independent. Thus, the processor knows that they can all be dispatched simultaneously without even checking. In present superscalar architectures, multiple instructions can be dispatched in parallel, but the CPU has to have a lot of logic to check and ensure that instructions are independent before they can be dispatched. You see, VLIW moves the independence-checking logic out into the compiler much as RISC moved instruction-decode logic out into the compiler.

    Those poor compiler writers. Oh well, I guess it keeps them in business :)

  • Time to start pumping more money into my AMD stocks... Intel is going down. I think the Athlon was the pivotal point where intel will fall completely to AMD...

    -- "Microsoft can never die! They make the best damn joysticks around!"
  • Since when has halloween been a major target for product release. The way the article is written, it's like these companies are upset because their going to miss some major seasonal spending events. This just sounds like your average production delays. Intel try to come across all clever and ambitiously post quick availability. Intel run into 'completely unexpected problems that they would never normally have to provision for'. Intel gets more egg on its face.

    It's not even as if the share price is going down because their going to make a loss this quarter. It's going down because the expected profit is LESS HIGH than expected.

    All in all, it's not that serious, some people are just a little unhappy that there have been delays. It's the armagedon that the article implies!


  • Ok, This is coming from someone on a Unix/Linux (majority) web site?

    No. /. crossed the barrier to a majority of readers running Windows over a year and a half ago, IIRC.

    Steven E. Ehrbar
  • IIRC, Word has been multi-threaded since version 6 (circa 1995). Admittedly at that stage it was for 'background printing' but still...
  • by mybecq ( 131456 ) on Friday September 29, 2000 @04:41AM (#745008)
    A bulletin earlier today contained unconfirmed reports of a small rebel chip manufacturer who has just finished shipping a product on time.

    Industry analysts were stunned for several hours while the small manufacturer's share price rose sharply.

    Details later revealed this company to be in the business of potato chip manufacture. They had just released their quick-double-dip-chip, widely accepted as the major driving force in the development of the cutting-edge rip-n-quick-n-dip-n-lick-n-chip technology.

  • Itanium may run IA-32 applications, but it does so slower than IA-32 processors such as Pentium or Athlon, since that's not it's native instruction set. Sure, the translation is done in Silicon, but nonetheless it's still there, and the effect on performance is apparent.

    AMD's Sledgehammer will run IA-32 applications FASTER than any current IA-32 processors, and therefore seems to be a true upgrade. I'd regard Itanium as a downgrade unless the OS and applications I wanted to run had all been recompiled/rewritten to run native.
  • by Markvs ( 17298 ) on Friday September 29, 2000 @04:44AM (#745010) Journal
    The P4 chip isn't that big a deal, but the 400mhz bus will speed things up a heck of a lot above my 133mhz bus. The processor speed is almost superfluous at this point.

    As for production work, be young have fun & buy Alpha. Four out of five SQL administrators whom have tried Alpha recommend it to their pat... er, users. :-)
  • I ask this in advance because the slashdot crowd usually has cash to burn.
  • If you are running a PII 233 that should be on the intel LX chipset that has a FSB at 66mhz. Even if you were lucky enough to have a BX chipset, i bet its limited to a 100 mhz FSB and not overclockable to 133.

    If you have a LX you can run a PII 333 or any Celery up to 600(its the max i think). Even if you had a BX, you can only go up to 700 mhz (ONLY i love it)

    If you are going to go after the 1ghz PIII, you are going to have to buy a new motherboard anyways. You might as well buy a High quality abit or asus KT133. But why not wait 3 months and get the AMD 760 DDR chipset. But hey, the SMP + DDR chipsets are coming out in January. Get one of those!!!

    dual 1.5ghz corvette (palmetto) w/ 133 ddr goodness. Drool!

  • And you won't get the *2 speed boost, more like 1.2-1.6 due to the fundamental difference in design.

    Remember that Intel pitched a P!!!-750 against a P4-1.5GHz in their initial demo...
  • Why would anyone need an SMP word? I mean even a puny 8086 can handle the speed which even the best typists type.
  • Actually, I remember that being Dan Quayle... gonna go look that up...

    Yep, it's Dan Quayle [].

  • I do belive that Word is a reference to M$ Bloatware Word. That piece of crap I can outtype on my Office P-2 400. And I'm no fast typing secretary...
  • Main processor speed is becoming less of an issue with specialized graphics processors picking up the entire processing load. Intel lost that battle when the GEForce came out. Although they continue to try to integrate graphics into X86 chips, it's a moot point.
  • is a Pentium 4, a Pentium 4, a Pentium 4.
    Gee, if I could only have a Pentium 4, then I could wish you "Merry Christmas."

    It seems so long since I could say, "Faster Faster Faster Faster Faster." Gosh, oh gee, how happy I'd be if I could have a P4.

    All I want for Christmas is a Pentium 4, a Pentium 4, a Pentium 4.
    Gee, if I could only have a Pentium 4, then I could wish you "Merry Christmas."

    Now if there's an EMBED tag...
    dd if=/dev/random of=~/.ssh/authorized_keys bs=1 count=1024
  • Frankly, I doubt very much that system makers are THAT unhappy about the P4 release and Christmas.

    Considers this:
    the P4 is going to cost MUCHO DINERO when it will be released, so who will buy this at the beginning?
    Answer: mostly companies which needs CPU power (CAD,etc).

    And you know what?
    Companies don't buy PC for Christmas!

    I think that it is more a case of journalistic exageration than a real problem..
  • by Anonymous Coward
    Personally, I'm waiting for Hemos to tell us the pentium 4 is Delayed.
  • RISC doesn't mean fewer instructions it mostly implies the instructions are simpler (don't do memory loads before addition followed by a memory write). Also RISC processor are more register based than memory based; in other words operations happen on registers on the CPU and the result go back into registers on the CPU.

    This allows simplier op code dispatch logic and simpler functional units.

    As you stated all of the major CISC CPUs have a more RISC like core that is fed by a decoder that splits legicy CISC instructions into more RISC like ops (or more likely microops).
  • ...unless you turn on spell- and grammar checking.


  • Godot called. He said he'd be late

    That is possibly the funnies .signature I've ever seen on SlashDot.

    Hm... NecroPuppy.. You a Skinny Puppy fan by any chance? :)

  • contrary to what microsoft would like to make you think, Windows 95 and 98 are more predominatly used in many companies than NT and 2000. And an extra processor in one of those machines is dead weight.
  • "That's hard to swallow," said a source at one PC maker.
    They're obviously still working out the bugs in the Cool Ranch flavoring.

  • I would go with an AMD over the Intel anyway. AMD is supposedly getting the Thunderbird (Athalon in a socket instead of a slot) ready for multiple CPU Motherboards and they are just as powerful and they cost about 1/4 less (at 1 GHz right now). AMD's plant in Dresden Germany and there other main plant in the states are not haveing any problems with supply and the quality of the Thunderbird I just bought is EXCELENT! I say forget intel and just get an AMD.
  • Apparently the system makers aren't that happy...

    Nobody with a rational thought in their puny heads would buy these 1.138745 GHz CPUs and Pentium-4's (5-4). Not in business, not in the home. You pay ALOT for something that'll be half the price (for the home market) next year. Not to mention how many bugs and fixes you'll get pushed on. Strangely as it seems, the most buggy shit seems to be most expensive and also the least decreasing in price over time. We live in funny times..

    Oh, THEY're happy. They're happy they get media attention. WE're stupid to let us bother with such nonsense.

    - Steeltoe
  • by Black Parrot ( 19622 ) on Friday September 29, 2000 @04:30AM (#745033)
    > Vapour P4 to avoid people buying AMD processors ?

    But is it working? I noticed a 1G Athlon system on the shelf at Best Buy yesterday. And the price wasn't unreasonable, considering what people normally pay for PCs.

  • by grahamsz ( 150076 ) on Friday September 29, 2000 @04:52AM (#745037) Homepage Journal
    Well this isn't really true.

    True most people will only have one active window at a time... you cant have any more on windows. But look at the amount of crap running in the background. My flatmates twin celeron feels VERY responsive because even when one app is hogging a cpu it's got one free to service all the routine crap.

    Word itself has been threaded for donkeys years, and any multithreaded app should automatically become an SMP version when there is more than one cpu. Also given the state of windows programming i'm starting to find the strong benifits in multithreading shit just from a coding point of view.... debugging is a whole other matter though.

    Additionally as SMP becomes more accepted we will soon see more and more supporting it, since most software developers out there have some deep resentment cauasing them to want to strike down intel and micro$oft.

    The other point is that word itself SHOULD NOT be able to occupy a whole p3 700 cpu for any length of time anyway... it's just an abuse of resources and we all know that microsoft wouldn't stand for it.

    Bear in mind that AMD hope to have dual cpu duron and athalon mobos out in the next 4 months, and have 4 and 8 way ath boards out by the end of next year.

    That should shake things up a bit, particualrly if you could run 8 $100 duron 1.5ghz (sounds about right for the end of next year) for little more than intels latest and greatest 2.6ghz p4.
  • Did anyone really expect this to come out on time. Kinda makes me think or Merce...err Itanium.

    Some people are predicting th death of chipzilla... I don't see that. Just quite a few people moving away for a while. Which is good for the whole industry.

    On a completely different rant...What is with all these damn companies and their vaporware? Seriuosly. Yopy, Itanium, SMP Athlons, various linux based webpads, a cell phone that works with a palm, a decent affordable mp3 player... It's rediculous. I read about this terribly cool stuff everyday but there is no possible way to get it. Ack!

  • Holding up the horrific 16 to 32-bit transition debacle (as executed by Microsoft with Win9x) as a good idea seems a little odd, considering it's been 13 years since the 80386, and most users are still crunching 16-bit code on their PIIIs and K7s.

    But, that's exactly what Sledgehammer is going to get you. No "64-bit" OSes (except for maybe Linux), but instead a bunch of small incremental "Accelerated for Sledgehammer" drivers and video games. And like, the 640K barrier before it, it's no real solution to the upcoming 4GB barrier ("ought to be enough for anyone", right?), which is the main reason you want a 64-bit chip to begin with.

    My guess is that Intel learned their lesson from the not-yet-complete IA-32 transition, and wanted to put in small disincentives that would hurry the transition to IA-64. That and marketing Itanium OS support like hell to all major providers, including Sun, IBM, and DEC (although they all reconsidered and said no), as well as funding Linux development.
  • by slothbait ( 2922 ) on Friday September 29, 2000 @05:58AM (#745048)
    ...that Intel chips are RISC on the inside as well. No one has built a truly CISC chip in years and years. RISC won the architecture war, but not the ISA (instruction set) war. x86 was too firmly entrenched. Thus we are left with modern architectures that emulate old, crufty ones. Not conceptually lovely, but functional enough.

    In practice even chips like the PowerPC aren't really RISC processors in the classical sense -- they implement too many instructions. (Altivec, anyone?) They merely hold onto the Load / Store memory model and the general feeling that instructions should be short and sweet. But they are far more complex than the RISC designs that academics came up with.

    A lot of students will take an undergrad computer architecture class and come away with a RISC chip on their shoulder. Plus, it's fashionable to hate Intel, and to rag on x86. So you hear a lot of "RISC rules, dude!". However, it's all a little silly. Internally, modern x86's have benefitted from all the advances of RISC design. All we are left with is the external interface from the old days. But how much does that really matter? Virtually no one writes inline ASM these days. If your only interface to the processer is through a C compiler, then you're never dealing with the ISA anyway.

    The story of x86's life: not lovely, but quite functional.

    just a few thoughts...
  • It's Itanium (aka Merced) that's predictably inching closer to cancellation, not McKinley which has only just taped out. Intel's IA-64 architecture was a joint Intel/HP effort, but the Merced/Itanium implementation was an Intel only design. McKinley is a complete redesign by HP, and AFAIK is expected to meet it's design goals.

    I agree that AMD's Hammer looks better positioned though (mostly due to being IA-32 compatible vs IA-64,s software emulation).
  • Actually Word would fair quite nicely on an SMP system. It is quite heavily threaded, and likes to do stuff like spelling and grammar checking in the background.

    I disabled this feature on my laptop because it took about 45 minutes off my battery life when using Word. A thread that can consumed 50% of my battery in an hour most certainly can benefit from another CPU.

  • Threading is evil and almost always more trouble than it's worth. You can usually (not always) get better results out of intelligent event handling or seperate processes than you can get out of threading a single process.

    It's a quick fix solution with nasty side effects.

  • That subject aughta tell it all. Even though the anti-overclocking efforts of AMD make me a little restless, what we may see here is a blessed occurrence of a monolopy break-down. If AMD are able to play their cards right and release a Pentium 4 competitor before the Pentium 4 is even released, manufacturers will flock to them in droves. That event would have serious and very positive reprecussions throughout the entire industry. I'm sure first and foremost, the notion of IntelM$ comes to mind. Come on AMD. The ball is in your court. Run with it! Kick Intel while they're down.
  • > Since when has halloween been a major target for product release.

    Considering the scary flaws, licensing terms, and privacy violations endemic in recent products, I think Halloween is an altogether appropriate time to buy computer stuff. I'm going to put one on my porch to scare off the trick-r-treaters.


Don't tell me how hard you work. Tell me how much you get done. -- James J. Ling