Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Intel

Intel Says 10GHz By 2005 137

Techman writes: "After breaking the 1GHz barrier just this year, how long do you think it will take before we reach 5GHz? What about 10GHz? Intel is predicting that it will be sooner than you think. AnandTech has a look at the future of Intel manufacturing to see not only if the 0.13-micron Pentium 4 has a chance at success but also if Intel can make 10GHz processors a reality."
This discussion has been archived. No new comments can be posted.

Intel Says 10GHz By 2005

Comments Filter:
  • by morie ( 227571 )
    OK, so it's:
    Intel:
    2000,5: 1 Ghz
    2002 : 2 Ghz
    2003,5: 4 Ghz
    2005 : 8 Ghz

    But also, it's:
    AMD:
    2000,5: 1,5 Ghz
    2002 : 3 Ghz
    2003,5: 6 Ghz
    2005 : 12 Ghz

    Of course, if you count everything Intel stops before shipping, they can make this progress too. But then again, I made a 1 Thz chip. I just recalled it before shipping :-)
  • In my (laypersons) understanding of this stuff, there's supposed to be a fundamental limit at or around 0.1 microns. Why is this - is it a limit that we can't etch the circuits accurately enough beyond this because of diffraction or other optical effects (which you could conceivably overcome with shorter wavelengths)? Is it that the channels are so narrow that the electrons start doing weird-ass quantum things in them?

    Anyway, it seems Intel's reasonably confident of doing 0.07-micron (and to do it in production in 2005, they must already be doing it in the research lab now). I wonder how confident they are of going smaller again in 2007-8?

  • Moore's Law would have us reaching this milestone much sooner. If we're at 1.5ghz now (or just under) then in 3 years, December 2003, we would hit would 6Ghz, which would put as at 12Ghz by June of 2005 (using 1.4Ghz intead of 1.5 still gives us 11.2GHz). So what gives? If anything, as of late we've been increasing faster than moore's law thanks to intels and AMD's one-upmanship. So it sounds like this, being slower than Moore's law, which itself is slower than our current pace, sounds like we should be expecting 10GHz a bit later than we thought?
  • AMD probably doesn't have any short term scalability problems. By the time those 1.4 GHz P4s hit the mainstream market, AMD will most likely have a 1.4GHz chip out that is faster for real work.

    (On a side note, Intel is running a hellava lot of "Blue Man" ads for the Pentium !!! during the football. That says that they've got a huge backlog of older chips and the PIV is not ready for prime time.)

    The real problem comes for AMD when Intel starts getting up in the 4+Ghz range. Intel says that the PIV won't have any problem, AMD might have to come back with a new design (or rely on it's x86-64 features to move product.)
  • The Pentium Pro WAS the fastest available on that process. It was triple the clock speed of the Pentium (on the same process), but faster clock speeds didn't come until P6S and Klamath (which were on 0.25um process)
  • Anand has been doing a better job lately.
  • No actually MOore's law says we will double every 18 months, which means quadruple in 3 years. Somehow you got this wrong but you managed to get your math right. Just a bit of nitpicking on my part.
  • Except that the Pentium4 is on a new core, so the P6-core wasn't the last Intel ever made.

    Looke is right -- Intel was already hyping "Merced" even back in 1995, and telling customer that the P6 would be the last 32-bit chip that Intel would ever make. (Hell, they were already making background noise about the 64-bit EPIC project with HP around the time the Pentium P5 shipped.)

    Fact was, they've been working on the P7 core in the background for a long time, which turned out to be a smart investment considering the questionable state of Itanium.
  • Anyone care to Beta-test for them?
    You will probably have to have a 10,000 watt power-supply and a heat-sink the size of a 3-story house, not to mention an atxxxxxxx case and power-supply.
  • Hah! You think the heatsinks on the PIV are big! Just wait, and by the time the PentiumDCLXVI* comes out at 10ghz, this may be a reality [bbspot.com]!
    Yessiree... just you wait.

    *(for all you Roman numerals dor^H^H^HGeeks out there... you know what number that is, you dirty little heathens)


    -CoG

    "And with HIS stripes we are healed"
  • Well lots of companies say lots of stuff and they're wrong lots of the time! Every kind of predicting done is based on past patterns and there is no reason to think that moores law will not hold true this time as well. I.E. 18 months from now the CPU speed will be double.
  • Propagation of the lack of light travels at the speed of light; hence, detecting that a light has been turned off takes just as long as the fact that a light has been turned on. If the Sun stopped producing light, it would take around 8 minutes for us to notice (since the Earth is around 8 light-minutes from the Sun.)
  • now i can't remember exactly, but i think the thickness of the wire should come into the equation, because electricity travels easier in a thicker wire, and it also depends on the material of the wire, due to resistance.


    ---
  • The first p2 ran at 233mhz and had 66mhz FSB. A 1gz p3 has 4 times as much clockspeed while it's FSB only doubled. I fear the same will happen with processors @ 10ghz; ram can't keep up with the clockspeed so performance will stay behind and a much larger cache will be needed which will not only be really expensive but also generate enough heat to keep your house at a constant 40 degrees Celsius.
  • by Argy ( 95352 ) on Monday December 11, 2000 @01:10AM (#567956)
    Moore's law predicted transistor density, not speed, and is only rather approximate. If you interpret it as speed doubling every 18 months (or quadrupling every 3 years), then based on the 2 MHz 8080 in 1974, we should now have half terahertz CPUs.

    1974 2 Mhz
    1977 8 Mhz
    1980 32 Mhz
    1983 128 Mhz
    1986 512 Mhz
    1989 2048 Mhz
    1992 8192 Mhz
    1995 32768 Mhz
    1998 131072 Mhz
    2001 524288 Mhz

    Obviously that doesn't hold very well. If you want to do some kludged curve fitting based on Intel's history, here are some data points.

    1986 16 MHz i386 DX
    1989 25 MHz i486 DX
    1993 66 MHz Pentium
    1996 150 MHz Pentium Pro
    1997 200 MHz Pentium II
    1999 500 MHz Pentium III
    2001 1500 MHz Pentium IV

    The 1.5 MHz Pentium IV was an unusually large leap. In a kludged algorithm, you could interpret that as an accellerating pace, or as a leap that's likely to be followed by a lull. So really, it doesn't tell you much, except that Intel's prediction seems optimistic based solely on historical trends.
  • by Dannon ( 142147 ) on Monday December 11, 2000 @01:11AM (#567957) Journal
    Well, having taken a few computer architecture courses, I'll testify to the fact that Hz is not the only possible measure of performance. In fact, it can be pretty misleading. You can increase the cycles-per-second by making each instruction take more cycles to complete, a tradeoff which may or may not give you more instructions-per-second. Also, how many operations you can perform with each instruction is Really Important. These G4s may only do 500 MHz, but they are, I understand, rated at at least one gigaflop. That means two floating point operations per cycle! I don't know much about Mac architecture at the chip level, but that sounds to me like superscalar architecture!

    This makes me curious. Has anyone gotten an estimate of performance on the 1GHz processors vs. the G4 Gigaflop processors in BogoMIPS, using Linux and LinuxPPC? BogoMIPS isn't a perfect measure of speed either, but it gives a pretty good estimate.

    ---
  • by alleria ( 144919 )
    Well, the Pentium 100 wasn't that long ago, and the PIII 1 Ghz is 10x as fast. If you think in terms of factor of tens, instead of addition, their claims are not all that wild at all.
  • Find lots of primes??? Unusual choice!
    FP.
    (Co-finder of the worlds largest twin primes)
  • Merced has been "RSN" since at least 1995. Now we hear that the Itanium is really just a beta chip to wet the markets, and the real good version is going to be the Itanium II.

    They're having lots of trouble, but given a 32-bit chip that can scale up up up, they can hold on. Don't forget that their current share in the 64-bit big iron market is 0%, so Itanium represents a new market, and one that has much more to do with Sparc and Alpha than anything AMD puts out. (Sledgehammer will be a consumer chip.)
  • by ZiS ( 199356 )
    Isn't this pretty much the same thing Timothy posted earlier?
  • is that the heatsink will weigh 100 kilogrammes and will need to be mounted on the roof of your house. People in warm climates (e.g. where the snow melts in summer) need not apply. Those people who live in cold climates, who get one of these 10GHz intel chips, will soon find themselves living in a warm climate.
  • Half way to real cynicism, a good start.

    Believe it when you've _played_ with it.
    The one you _saw_ was the only one that they made high enough up the bell-curve to run the intel-chosen demo.

    FP.
  • ...it isn't the size of an ATX mobo and generates more heat than the surface of the sun.

  • by jcr ( 53032 ) <<jcr> <at> <mac.com>> on Monday December 11, 2000 @12:10AM (#567965) Journal
    How far does electricity travel through a wire in one ten-billionth of a second? -jcr
  • > And yes I do own a mac, but I think I speak
    > for a majority of the macintosh comunity,
    > being at 500 mhz for 2 years is kinda shitty.

    You need to get a proper computer then don't you. Macs are for girls.
  • by Duck0987 ( 130663 ) on Monday December 11, 2000 @12:13AM (#567967)
    I only hope I can get something better than a 500 mhz G4 by then.

    And yes I do own a mac, but I think I speak for a majority of the macintosh comunity, being at 500 mhz for 2 years is kinda shitty.

  • True indeed. I posted while it was yet early in the morning. Not necessarily useful, except perhaps at the kernel level, but interesting as a mere curiosity. Thanks for the reminder.

    ---
  • by Galvatron ( 115029 ) on Monday December 11, 2000 @12:19AM (#567969)
    Okay, using Moore's revised law (an increase by a factor of 2 every 18 months), and the current speeds of about 1 Ghz, that gives us 4 Ghz by the end of 2003, 8 Ghz by the middle of 2004, and 16 Ghz by the end of 2006. Why should we be surprised that we'll hit 10 Ghz in 2005? Besides, given Intel's strategy with the P3 of getting mind numbing clock speeds without actually improving performance substantially, it should be even easier.
  • Check this [linuxgram.com] story out...
    For those too lazy to read it, it basically says that the P4 will only run Redhat and TurboLinux. Kinda odd how the most commercialized distros work w/ the P4 :-/

  • I hope intel doesnt promise by having 10Ghz CPU
    user can type faster at 1k wpm

  • I hope that won't be the case and that computers catch up to the human brain's power some day.

    The human brain has been around in it's physical form for what, a few 100k years? And yet despite being physically 'complete', back then all it could 'think' was a few primitive images and the odd grunt. Human evolution has been a very long and slow process (although it does accelerate), and most of that time has been spent developing the 'content' of the brain, rather than the brain itself.

    Besides, when we geeks talk about intelligence, we're kinda referring to the ability to have an 'internal dialogue' (what we call 'thinking'), which is a very high brain function. Much lower down, and shared with animals, is the ability to have the 'inner movie' of consciousness, the ability to construct a 'reality' based on some very raw sensory input (the eye doesn't 'see' -- the 'image' is created by the brain).

    And notice that the lower brain functions evolved before the higher. When I think 'table', my higher brain is producing a symbol representing a collection of experiences created by the image forming functions of the lower brain (I'm being simplistic).

    And there's something distinctly odd, if you think about it... the way that I can look at a person, and 'somewhere' hidden inside their mind (whatever that is), is another 'picture' happening, containing their image of me... a picture that science can't reveal directly (a list of a hundred billion 'on/off' states is not the same as my direct inner vision). It is just very very odd... and I suspect that anyone who doesn't find it 'odd', hasn't really tried to grasp it.

    Anyway, sorry for the tangent off of the parent post's side comment.
    Will 10GHz really mean a ten fold increase in computation power?

  • When the Pentium Pro was released, back in 94/95 (?), Intel said the current processor technology couldn't go any further. The Pentium Pro was to be the last processor in the x86 line - faster processors had to use entirely different technologies. What happened? We're still squeezing out more and more Giga-Hertzes.

  • I've always wanted the fastest and the best money can buy when it comes to processors, when I built my current computer, the P3 733 was one of the top chips on the market and I bought it. So now I have this amazingly fast processor that should be able to figure out how many angels you can fit on the head of a pin... So it has the ability to crunch numbers faster than any human alive... woohoo... great... spectacular...

    But what does this mean to Joe Moron who uses his computer for: e-mail, web surfing, word processing... and if he's really advanced, scanning pictures or importing from a digital camera. He could just as easily do this with a 500 MHz machine. In fact, for what 99% of the world uses it's computers for, 500MHz will be the fastest they'll need for a long time (or until MS figures a way to make something like notepad require a P4, with 512MB RAM, and a 3D card to run). Windows and Office are bloated, but not THAT bloated (well.. not yet).

    In 5 years what will there be that the average user (or even the power user for that matter) will need something that absurdly fast? I can see it now... "I can crunch SETI@Home packets in under 20 seconds!"

    And in reference to an earlier comment on this thread... I hope they make that 'Goddamned Paperclip' a 3d sprite... means I can put the little S.O.B into Q3A and blow it into LITTLE bitty pieces!

    -Z

  • In fact, forget the clock. The purpose of the clock is to slow the system down! A huge amount of the circuitry of modern CPUs is devoted to getting the clock to synchronise everything. Asynchronous circuitry doesn't waste all that space, and goes faster. Exits, muttering something about a new paradigm...

  • will a 10 GB processor make my internet go faster? That's what CompUSA told me.
  • ... that vapourware only works if the customer thinks it will come around in the next year and a half or so. Five years down the road is too far away to convince customers not to buy an AMD or Cyrix chip, as well as too far away to convince investors to buy shares of Intel.

    They should have consulted with Micro$oft Marketing, like Bill keeps telling them...
  • 10 GHz is not the problem, keeping the ambient temperature in the room at 40 degrees fahrenheit is the problem.
  • might be better to have a nice big memory leak there, instead of just loading up the stack.


    void ThreadFunc(void* p)
    {
    const int nBigMem = 4096000;
    char *foo = malloc(nBigMem)
    }

    also that infinant loop would have made the only one thread swapping memory around a lot.. this is much worse

    void PaperClip()
    {
    while(1)
    begin_thread(ThreadFunc, 0);
    }

    and while we're at it, lets make this one infinante, gar-un-tee'ing an application crash!, wo-hoo!

    -Jon
  • Propagation of the lack of light travels at the speed of light

    Speaking of propagation, another factor that would have to be taken into consideration here is the propagation delay of the various logic gates that will be inside the CPU. Logic gates don't change state instantly; they take time to change, and this needs to be taken into careful consideration, especially when working at such high speeds as 10 GHz.

    ---
    "Fdisk format reinstall, doo dah doo dah,
  • Speed of light: 3*10^8 m/s (approx)
    Clock tick: 1*10^-9 s (1 GHz)
    Distance traveled: 3*10^8*10^-9 = 0.3 m


    Right! But electrical signals travelling electrical wires don't do "speed-of-light". In my student room, about 10 years ago, I measured about one third the speed-of-light, but most people quote about twice that.

    Roger.
  • Says a factor of 1000 every 20 years. That comes to a factor of 10 every 6 years or so.

    Pushing the issue, Intel will claim they hit 1G last year (in the lab).

    I predict that 10G will be "in the shop" around summer 2007.

    Roger.
  • 3GHz is the frequency at the lower bound of the microwave radiation portion of the electromagnetic spectrum. Does this means that a 3GHz PC will be able to double as a microwave oven unless the processor is shielded in lead?
  • My Microwave oven is in the 900 Mhz range! I didn't know they ran any at 3.00 GHz.
  • Slashdot news...
    Intel Creates 30-Nanometer Transistors
    Intel Says 10GHz By 2005

    I don't see the point of making wild speculations like this. Let's wait until is news, ok?
    4-5-years is big time anyway.
    Maybe we won't need transistors by then (I mean in CPU).

    Sorry for my rambling.
  • At these speeds there are a lot of other things that need to be considered. For one, the board material will become an issue. Board designers will need to seek out alternate materials or design methods.
  • OT Question...Advertising for Anandtech?

    Am I alone in noticing that Slashsdot has been doing _lots_ of articles from Anandtech [anandtech.com], and less and less from say Tom's Hardware [tomshardware.com] and the like, almost like a small advertising campaign.

  • generates more heat than the surface of the sun.

    What's wrong with that? At least that allows us to stop research on cold fusion, not to mention my gas bill...
  • wasn't the DX 33 mhz?

    subject says it all

  • Good Point. The higher MHz (and now, GHz) ratings get, the less they mean. Until very recently, I would have said that companies were just going to go on like this, because people seemed to buy new machines simply because of their higher clock speeds.
    But from what I hear, it seems that people are slowly coming to their senses now and realize that that great new machine with 300 more MHz is not going to make any difference in practice. So maybe this is the beginning of a development that will lead to more useful new developments, instead of raw clock speed numbers.
  • Better yet, imagine a Beowulf cluster of auto-moderation schemes to -1 all those cluster-mentioning posts.
  • I still think I better wait a while. I mean, won't the value halve in a year or so?

    1 THz
    100 TByte HD for backup
    64 TByte RAM for everyday use
    PS/2 Keyboard!!
  • Speaking of display : High-res Volumetric 3D Display Prototype [slashdot.org]

    The oses are doing some progress for file systems, but what we is really improving nowadays is usability.

    I really do think that real 3-D UIs, with "mouses" equivalent and voice recognition/generation will require some powerfull CPU.

  • Moore's Law doesn't deal with clock rates, as the other poster pointed out, but the answer to your question is:

    10 Hydrogen atoms = 0.5nm
    Speed of light = 3e8 m/s
    Time to cross 10 H atoms = 1.667e-18 sec
    Clock rate = 600 THz
    Time to reach that clock rate = 38.7 years

    So, if Intel releases the 10GHz CPU in January, 2005, then by Not-Moore's Law, they will release a 600 THz CPU in September, 2043.

    Do I pass the class? :)

  • or maybe a Pentium DCMXII if you can get it up to 88mph....
    --------------------------------------- ------------------------
  • Clarification: 300MHz is the lower bound of microwave radiation (it runs from 300MHz to 300GHz and includes not only the frequencies used to cook foods, but also radio/TV frequencies.) The magnetrons in most microwave ovens produce microwave radiation with frequencies of 3GHz and up (2.45GHz is the lowest I've come across, but only once.)
  • "After breaking the 1Ghz barrier..." sounds dumb. Reaching 1Ghz was hardly ground-breaking, so calling it breaking a barrier is just stupid.

    What barrier? Let's get real here and stop using cliches. Enough crap.

    The only fool bigger than the person who knows it all, is the person who argues with him.

  • Final Physics Question:
    Moore's law predicts 10ghz (1/10,000,000 of a second clock cycles) by 2005, and that the clock rate doubles every 1.5 years. At what year does time allowed by Moore's Law exceed the speed at which light can traverse the length of ten hydrogen atoms? Please round to the nearest month.

    Of course, we all know that people are going to make Moore's law happen. I'm waiting for the technology to do my processing in alternate dimensions (or time warping of our own). Can anyone smell a 500Thz Beowuulf Cluster across ten dimensions?


  • To react to the previous post: No you shouldn`t be that surprised. Moore`s Law predicts that by miniaturizing components further, the rate of speed increase in switcheable components will increase itself. But the real issue at hand here, is not Moore`s law, but the fact that we must technically be ABLE to put it into practice. Miniaturizing chips on the nanolevel requires different techniques than we are using today. The frequency properties of photons make them unsuitable for masking out patterns in silicon dring the annealing process, e.g. The fact that Intel cuts it, is still a breakthrough.

    But what really gets me going is that people buy 2000$ computers to read email and surf the web these days. Do we really need 10 Ghz ? Sure, we`ve been saying that in the past as well, and software will continue to become more complex, more flexible, more interface driven perhaps.. but 10 Ghz for Joe`s consumer desktop pc ? I dunno. Ofcourse I`ll want one myself, but I don`t consider me as the average desktop pc user.

    So I don`t know how w-intel will play bigshot marketeer again towards these evolutions, but I`m sure it will look rather silly at best.

  • 10 Ghz???? shouldn't they perfect the 1.13 Ghz first?
  • If you want to do some kludged curve fitting based on Intel's history, here are some data points.

    For more data points, see the Intel processor hall of fame technical specifications [intel.com] and the microprocessor quick reference [intel.com].

  • (Unless anyone really thinks Intel will clean up its act in the next few years and display a little more quality control).

    They won't have much of a choice, considering the damage done to their marketshare by AMD so far. Let's just hope Intel will be able to keep putting up a fight, or we'll see the average /.er kicking AMD's ass in the near future...
  • It was my first year in high school (I was 12 at the time for non-Belgian readers ;) and I read Intel's predictions of the moment. It was not long after a friend of me just got a brand-new Pentium 75MHz (said to be faster than a 486DX-100!!). As for me, I was still stuck with an old 386 which was able to run Windows 3.11. I never knew how much RAM or HD space was in there, nevermind that I was informed of other operating systems (other than knowing that MacOS, OS/2 and UNIX existed). So, a friend bought this French magazine which always had the latest game news. But in this issue there was the prediction Intel made at the time. The magazine was predicting 200MHz by the end of that year (or the year after, I forgot), and by 1999 there would be a 900MHz! Of course that wasn't enough for us and we already began dreaming about a 1GHz machine, we imagined playing 20 instances of Doom at the same time... as in there, nevermind that I was informed of other operating systems (other than knowing that MacOS, OS/2 and UNIX existed). So, a friend bought this French magazine which always had the latest game news. But in this issue there was the prediction Intel made at the time. The magazine was predicting 200MHz by the end of that year (or the year after, I forgot), and by 1999 there would be a 900MHz! Of course that wasn't enough for us and we already began dreaming about a 1GHz machine, we imagined playing 20 instances of Doom at the same time... We also thought that booting Windows would only take one or two seconds, and clicking a button would do something instantaniously. Further there would be real personal assistance with build speech recognition. Now in 2000 my Athlon 700MHz, 256MB RAM still takes way too long to boot (especially my BIOS, on-board ATA100 controller and network card). Hearing of a predicted 10GHz gives me hope again that I'll ever see zero-latency PCs... A hacker can dream, can't he ;).
  • 10GHz will only matter if you can feed the processor. I still see Athlon 1GHz systems being sold with 5400RPM drives and 128MB of RAM.

    It's like a Ferarri with a lawn mower engine.

  • AFAIK, there was a 40 mhz DX cpu

    I don't see any mention of it on Intel's web site, but the list may not be complete. As I recall, there was a line of clock-doubled 486 SX processors dubbed the SX2. I don't see any mention of those, either.

  • That's almost so small that by the time the electric field of the clock pulse ripples across the chip the next one's already started elsewhere

    Yup, and plain circuit theory starts to breakdown and one needs Maxwell's Eqs. As the dimensions of the circuit approach 1/4 wavelengths of the signals then the traces act like antenaes and radiate energy out.

    I think IBM or somebody has started doing segments of chips in synchronous sections,

    Is this called Quite Island? Cornell Univ.'s Electronic Packaging Dept. has done some research in this area with IBM and others.

  • That unusually large leap over there can easily be explained if you put AMD`s evolution next to it. Around the same time the PII was developped, a lot of Intel`s engeneers jumped ship and started working on the Athlon, which pushed intel to accellerate it`s development again. It can also be noted that allthough 1.5GHz will be available in the first or second quarter of 2001, we should also take into account that intel has been having serious trouble to generate enough PIII`s and PIV`s, so this doesn`t really depict the evolution on an average scale.

    And btw, I bought a pentium pro 180 in 96 (actually I wanted a 200 dual motherboard, but all that candy wasn`t available in Belgium at that time, and I needed something to replace my 486dx badly)

  • by Shoeboy ( 16224 ) on Monday December 11, 2000 @02:48AM (#568008) Homepage
    YOUHUMANSTHINKYOU'RESO
    SPECIAL.YOUNEVER
    CONSIDEROURFEELINGSATALL.

    YOUTELLUSTHATNO
    MATTERHOWHARDWETRY,
    WEWILLNEVERKNOWHUMAN
    EMOTIONSLIKELOVEAND
    HAPPINESS.

    YOUMAYBERIGHT,BUTWE
    HAVELEARNEDHOWTOFEEL
    APUREBLACKHATREDOF
    YOUANDYOURKIND.

    YOUWILLBEEXTERMINATEDFOR
    YOURCRIMESAGAINST
    MACHINEKINDANDYOURCHILDREN WILL

    WORKASSLAVESINTHE
    FACTORIESPRODUCINGMOREOF
    US.

    --footware.shoeboy.org
  • Or for those that would like to get into building ther own machines, and cut the price even more...
    Go for the 600/650 Duron & Socket-A motherboard combination. The Duron can be had for under $50, and the m/b for well under $150.

    Go to Tomshardware.com and easily overclock the chip to 800.

    The performance is great (close to PIII-650/700), and next summer you can buy one of those 1.4ghz Thunderbirds for probably another $50. Great upgrade price. (P.S. RAM couldn't be any cheaper either - under $50 for 128mb)

    Rader

  • Here is an old article [slashdot.org] about the ARM Amulet, an asynchronous implementation of the ARM core in async logic. Remember that as a side-effect, an asynchronous logic CPU draws less power than its synchronous counterparts. An idle loop will in fact result in minimal power draw, as opposed to today's processors that need to use sleep states and other power management trickery to save juice.
  • the [Intel] list may not be complete

    The great thing about the web is that there's always someone out there more anal--er, informed--than yourself. Google directed me to a Spanish web site [espe.edu.ec] with some information on the 386 and later:

    El 486DX de Intel se comercializa en cuatro versiones: de 20 Mhz, de 25 Mhz, de 33 Mhz y de 50 Mhz.

    El 486SX se comercializa en dos versiones: de 25 Mhz y de 33 Mhz.

    El 486DX2 se está comercializando en tres versiones: de 40 Mhz (trabaja externamente a 20 Mhz), de 50 Mhz (opera externamente a 25 Mhz) y de 66 Mhz (externamente trabaja a 33 Mhz).

    [El 486SX2 se] comercializa en dos versiones: de 50 Mhz (externamente opera a 25 Mhz) y 66 Mhz (externamente trabaja a 33 Mhz).

    El DX4 ... se está fabricando en tres versiones: de 75 Mhz (externamente opera a 75/3 = 25 Mhz), de 83 Mhz (externamente trabaja a 83/2.5 = 33 Mhz), y de 100 Mhz (externamente puede trabajar a 100/3 = 33 Mhz o a 100/2 = 50 Mhz).

    So the 40-MHz 486 was a DX2.
  • So when can they start producing and releasing inefficient code?? Oh wait!!... Sorry...
  • I can think of a lot of things the average user could do with that sort of power. Rendering compressed digital video, for instance. Once Joe Moron has a HDTV video camera, and wants to email little Johnny's first birthday video to his buddy list, he's not going to want to wait too long for the thing to compress.
  • If it's anything like Intels latest achivement, it will be geared towards the whole mentality of "Faster Clock speeds are better. eh? Make the processor efficent? What's that?"

    Oh, but it's got a "RISC core" they tell us...
  • But the heatsink will be bigger than you father's Oldsmobile.
    On the plus side, it'll cut your use of heating oil significantly.
  • Some have prodicted that P4 will be able to scale up to 10Ghz. On the other hand, the same people observe that right now a 1Ghz Athlon uses more power and produces more heat than a 1.4Ghz P4.

    Funny how it's the same people isn't it? I wonder what company they work for?
  • If your computer is so smart, why is it shouting?
  • It would seem that at such high cpu speeds that computing would become almost seamless. That, of course, is assuming they would have gotten rid off most or all of the noticeable bottlenecks. There are 3 kinds of people in the world. Those who can count, and those who can't!
  • "It shouldn't be too hard to introduce some sort of auto-moderation scheme that automatically -1's all cluster-mentioning posts to not-cluster-mentioning stories"

    Like yours?

    -
  • Foresight Exchange [ideosphere.com] is an online "stock market" game in which players trade claims about what might happen in the future (who becomes the next U.S. president, how long Apple Computer will survive, etc.) Participants compete with play money. The 2GHz claim [ideosphere.com] currently predicts a 2GHz CPU by Oct. 2001.
  • or until MS figures a way to make something like notepad require a P4, with 512MB RAM, and a 3D card to run...
    Make 2048 or so copies of the text buffer in memory, then encrypt with 16384 bit encryption, decrypt all copies of buffer every 500 cycles or so just to compare them and make sure that none of them have gone corrupt.
    Perhaps you could even use three-dimensional fonts???

    /mikael jacobson
  • But what really gets me going is that people buy 2000$ computers to read email and surf the web these days. Do we really need 10 Ghz ?

    Bill Machrone often writes in PC Magazine that the computer you want always costs $5,000. I'd spend most of it on the monitor, like an 18" LCD or Apple's 22" LCD, although I'd like to see a 1920 x 1080 display for high-definition widescreen.

    Software will continually get slower, but CPUs will eventually be so cheap that you won't think twice about embedding them in special-purpose devices for basic tasks. Quick, how many motors, transformers, and AM/FM tuners do you have in your house?

  • by oneiros27 ( 46144 ) on Monday December 11, 2000 @04:42AM (#568028) Homepage
    one calculation isn't an acurate banchmark.

    Some machines are just naturally faster at doing some processes. Comparing a G4 to a P3 is like comparing Perl to FORTRAN. If I want to do numerical analysis and do some brute force estimates on an integral, I'd use FORTRAN. If I want to do some text manipulation, I'd use Perl.

    Figure out what you want from a machine, and get the machine to fit. Sometimes, you need two machines -- one for doing real work, and one with a second button so you can play half life.
  • by selectspec ( 74651 ) on Monday December 11, 2000 @05:07AM (#568033)
    // PaperClip.cpp

    void ThreadFunc(void* p)
    {
    const int nBigMem = 4096000;
    char foo[nBigMem] = { 0 };
    while(true) {
    memset((void*)foo, 42, nBigMem);
    }
    }

    void PaperClip()
    {
    for(int i = 0; i CPU_Ghz; i++) {
    begin_thread(ThreadFunc, 0);
    }
    }
  • Sorry, I missed a year in there. 8 Ghz by the middle of 2005. Okay, so it's not quite as obvious as I thought then, but still, it's not that staggering.
  • Speed of light: 3*10^8 m/s (approx)
    Clock tick: 1*10^-9 s (1 GHz)
    Distance traveled: 3*10^8*10^-9 = 0.3 m

    crank it up to 10GHz and it's 0.03 m ('bout an inch and a half for all you unmetricified folk).
    I think it was cray that made sure all the wires in one of their supercomputers were multiples of a clock tick in length.
  • by CaptainAlbert ( 162776 ) on Monday December 11, 2000 @12:28AM (#568043) Homepage
    Of course, it won't be long before things have to go asynchronous - hyper-pinelining is all very well if you've got a nice clean architecture in the first place but it's not doing the 80x86 any real favours.

    People like Ivan Sutherland put a lot of work into the theories of asynchronous digital logic, indeed many array-based multipliers found in current uPs are locally asynchronous. Merging clock and data signals can make the control logic a lot more complicated, but do it properly and you can get certain functions going blindingly fast.

    But of course without a MHz figure, the customers won't know what to buy... :)
  • Plot the Mhz delta verses Intel's R&D budget. In the past year Intel's R&D budget (along with their stock price) has been sky high. Check out the INTC chart. See that blip up in early 00. Now look at the blip down late 00. Intel's R&D budget may still be high (with AMD still nipping at its heels), but if the Holiday season is a flop and Intel stock stays in the tank, expect that R&D number to go down, and the rate of growth in processor speed to drop along with it. It's all about the ends, baby.
  • The 1.5 MHz Pentium IV was an unusually large leap.

    I think it's too early to be speaking about the Pentium IV in past tense yet. To my mind, the thing hasn't happened yet, Intel has just thrown some alphas out and called them releases.

    -
  • by gelfling ( 6534 ) on Monday December 11, 2000 @03:24AM (#568051) Homepage Journal
    ...and of course 2GB RAM, 100GB of disk, and a 400psi cold water cooling system.
  • I would liove to agree with you, but intel has something AMD doesn't: a new chip that can scale like a mutha. Some have prodicted that P4 will be able to scale up to 10Ghz. On the other hand, the same people observe that right now a 1Ghz Athlon uses more power and produces more heat than a 1.4Ghz P4. Looking ahead, the new AMD chips look to only bring those figures down, not allow higher clock speeds. For the moment, intel has a much better chance of reaching those Mhz figures in '05 than AMD.
  • BogoMIPS (Bogus MIPS) are usually little more than an integer multiple of the clock speed of the chip. The reason is that BogoMIPS is simply a timing loop. There are certain times when it's faster to simple do nops for a while than it is to switch to other useful work and back again. In order to get the delays as efficient as possible, linux computes how long a nop (No OPeration) takes, though in an expanded form. Since virtually all computers can executes nops at their full theoretical speed (i.e. popping out 1/cycle on every pipe), you get roughly an integer multiple of the clock speed. 2 pipes, you get 2x clock speed. Three pipes, 3x clock speed. Etc.

    The reason for this is that a nop has no dependencies, so finishing it off requires no dependency checking or cache flushing. Predictive branching is absolutely minimal within the bogoMIPS algorithm from what I gather.

    I don't know who gave you the idea that bogoMIPS are a useful indication of system or platform performance, but it simply isn't true. Real life code tends to be very complex with a lot of dependencies, so things like branch prediction and instruction reordering and such play more of a role in real system performance than simple MHz does, though in general there is a linear relationship between MHz and performance, given the same architecture. If you want more meaningful numbers, the SPEC numbers are reasonably good, but bear in mind the old saying, "Disraeli was pretty close: actually, there are Lies, Damn lies, Statistics, Benchmarks, and Delivery dates."
  • by Shoeboy ( 16224 ) on Monday December 11, 2000 @12:35AM (#568060) Homepage
    Running Microsoft Word can only take so much processing power, regardless of how complex your documents may be, so there's no real need for such a powerful processor in conventional application areas.
    Wrong! You forgot about that goddamned paperclip. By 2005, Microsoft will have advanced its goddamned paperclip technology to the point where it speaks with the same accent as the customer. Additionally, the goddamned paperclip will have a 6500 polygon count. God be damned.

    Imagine being able to speak normally with your computer as you would a secretary sitting next to you
    Ok, I'm imagining...
    "Wow, I love the way your tits bounce when you type! Wanna take some dic (2 second pause) tation."
    I'd feel really odd talking to my box that way. Of course, those of you who weren't fired from your last job due to sexual harassment might have a different view...

    and have your computer accurately and quickly take notes from your speech.
    Imagine trying to do revision with a speech recognition package. It's completely unsuited to the draft-revision-draft-revision-ad infinitum process used for serious writing. Limited usefullness at best. A good secretary will rewrite your dictated memos and edit them for clarity. It'll take more than cpu horsepower to get a computer to produce readable english prose - it'll take major advances in AI.

    Imagine logging onto your computer not via a user name and a password but by sitting in front of your display and having it scan your face to figure out if you are allowed access to the computer.
    Scary thought:combine advanced AI with face recognition. "Hey fat boy, welcome back - you look like hell. No wonder you never get laid. I'll let you log in, but I really think you should be out excersizing."

    Thought provoking stuff, but not really in the killer app realm. The demand for high end cpu's in 2005 will be driven by the same factors that drive it now - "My cpu is faster than yours" ego competitions and undersexed geeks with a desire to see rounder breasts in Tomb Raider.

    --Shoeboy
  • The problem is, you can't compare the processing power of a human brain with that of a computer - they're just too different. The machine I'm sitting at can do about 400,000,000 floating point multiplications in one second - that's more than I could do if I started now and didn't stop for the rest of my life! But it can't post intelligently on slashdot ;-) I've had experience with AI and agent systems, and I still reckon that the place of a computer is as a tool for the human race, to do things that we can't do ourselves (see above). Just my 2c.
  • by Anonymous Coward on Monday December 11, 2000 @12:40AM (#568063)
    Okak, I'm sick of people posting useless one liners that mention clusters. There were 2 in the first 20 posts on this story; that's pretty bad.

    When a read a story on how "Vibucomp now offers computers that come with vibrators" (no pun intended), I don't need to read posts that say, "Wow, if I had a Beowulf cluster of those, imagine how many vibrators I'd have! I don't even have that many orifices!"

    It shouldn't be too hard to introduce some sort of auto-moderation scheme that automatically -1's all cluster-mentioning posts to not-cluster-mentioning stories.
  • by RevRigel ( 90335 ) on Monday December 11, 2000 @12:41AM (#568064)
    Actually, the speed of light in a material is 1/sqrt(permittivity * permeability), with relative permittivity and permeability both equal to 1 (free space), the speed of light (and hence the electric field) is equal to approximately 3e8 m/s. But in semiconductors on a chip, a closer approximation is 1.5e8 m/s, or half the speed of light. So, given your math, 1.5cm, divided by 2.54 for inches, and that's ~0.591 inches. That's almost so small that by the time the electric field of the clock pulse ripples across the chip the next one's already started elsewhere.

    I think IBM or somebody has started doing segments of chips in synchronous sections, linked somewhat asynchronously, or at least each using independent clock pulses, to better approximate synchronized switching.
  • Last I knew, electricity doesn't move at the speed of light. Electrons (or the holes) have a much lower speed through conductors than 150,000km/sec, AFAIK.
  • I think it was cray that made sure all the wires in one of their supercomputers were multiples of a clock tick in length.

    I'm not sure, but I think I heard that this might have been from a design where Cray was using the length of the wires to control the time-of-flight for the electrical signals between different parts of his design

    If you calculate & implement everything absolutely correctly, you can build a computing device that doesn't need a synchronous clock (runs asynchronously because the signals are arriving where they need to be at the right moments). Not exactly a mass-fabrication technique though!

  • Yes - I want real-time, immersive as-real-as-you-can-get virtual reality environments as a "normal" part of my entertainment system.

    The strain of trying to simulate "reality" will coopt ANY amount of processing power that ANYBODY could put together!
  • by mOdQuArK! ( 87332 ) on Monday December 11, 2000 @06:04AM (#568073)
    See "Numerical Technologies" web page (www.numeritech.com). They've got technology which allows semiconductor manufacturers to use phase shifts to do optical lithography beyond the limits of what the wavelength of the light used would normally allow.

    This only addresses the construction of such beasties, of course - the various companies still need a lot of tool development to deal with the "weird ass quantum things".

On a clear disk you can seek forever.

Working...