Forgot your password?
typodupeerror
News

Are High-End CPUs Worth The Money? 289

Posted by timothy
from the film-at-11 dept.
Rampaging Goatbert (aka Jeff Feld) has posted a story at Newsforge about something you may want to argue about with your boss or significant other. Specifically, whether high-end CPUs are worth their high prices. Personally, I look even lower on the processor food chain, but watching those price-curve inflection points makes the runner-up chips pretty tempting. Your mileage will almost certainly vary.
This discussion has been archived. No new comments can be posted.

Are High-End CPUs Worth The Money?

Comments Filter:
  • Buy a 386dx40 at $80, or a 486dx33 at $600?

    Buy a 486dx100 at $80, or a P1-166 at $600?

    Buy a P120 at $80, or a P233MMX at $600?

    Buy a P300-CelebA at $80, or a PII-450 at $600?

    Buy a Ath800 at $80, or a P4-1.3g at $600?

    You're always paying a premium, its the premium of boasting, and since the actual performance increase merits at max 50% (probably far less unless you're using max-cpu apps, and the rest of the kits not bottlenecks).

    So, No. Unless someone else is paying...

    Smid
  • This article is about the "high-end" of a low-performance architecture...

    When I saw the headline, I thought "At last, a real comparison of x86 v. UltraSparc, Mips, Alpha, et al. .."
    ... but no. Same old crap about Athlon and Pentium.

    Not even a mention the Itanium, let alone a date.

    Is the plural of Pentium "Pentia"?

    Is the plural of Itanium "Itania"?

  • CPU's except for the very latest top end are v.cheap now anyway. In all my Experience the BIG performance choke on Win NT/98/95/Linux/Unix is the hard disk/controller. old Pentium 233 much faster than PII 450 in all apps including gaming using 9Gb/7200rpm SCSI on DPT 64Mb Cached controller. Even on "cpu" intensive app's most of time spent waiting.... underclock your current pc by 25% and I'll be surprised if you can tell the difference.
  • barring technological leaps, incremental steps always cost more. it is also important to remember that the 1.33 is superceded by the 1.4, and back stock always sells for less that new production, so, in actuality you are getting a great deal on the 1.33 rather than paying a premium for the 1.4. also the article ignored the overclocking issue, it is possible that the 1.4 can be overclock a magnitude above the 1.33, especially if the 1.4 is an athalon 4, which runs significantly cooler (enough to allow it for use in laptops). on the other hand, i personally went with the athalon 850, with the first gen gigabyte 761 motherboard (which only supports only 200mgz memory bus, but then i had it a couple months before anyone else) and it screams and i have no regrets.
  • I'm getting a 1.4 Ghz AMD chip is so I can out do my friend who just bought a 1.33.

    It's like having a new Porsche compared to your neighbors new Camaro

  • Where I work, we have multiple boxes running Dual 1000ghz processors /w between 1-2 gigs of ram.. and they aren't powerful enough. Sure, we can build more machines.. and we do, we add at least 2 machines a day.. often more. For us, fast boxes are pretty essential.. for Joe Average Linux user.. he doesn't need much more then a pentiumII :)
  • This may have been more pertinent back when I paid $500 for the latest and greatest Intel Pentium II-400. Now with Intel and AMD duking it out in the "Great American PC Price War" it is more of a non-issue really. Back when Intel was the only reliable chip manufacturer (Yes I did try one of them Cyrixes) they could set any price they wanted, and the premium for the top of the heap cpu was ~$200 or more. Right after I spent $500 for my chip, every tech savy bargain hunter in the world bought a celeron 300a and overclocked it to 500mhz or greater. You could get more performance for significantly less money.

    Now the 1.4g Athlon I just bought (to replace my aging p2-400 which incidently goes for $58 now) cost just over $150. Sure I could have saved $30 and gone with the next one down, or even saved $80 and gotten a 1g. But come on, $80 only fills my gas tank 3 times.

    Now I totally understand if you are talking about spending $568 for a P4-1.8g, or $326 for a P4-1.7g, ($242 markup for .1g) or anything P4, for that matter. Damn those things are expensive. But if the rich gotta-have-its out there don't mind paying for it, thats fine with me. That's why they sell a wide range of cpu's instead of just one. The huge profits from the high end make up for the R&D and other overhead. Honestly how much do you think Intel makes from selling a Celeron 600 for $35?
  • google (Score:5, Insightful)

    by fjordboy (169716) on Monday August 06, 2001 @03:27PM (#2163743) Homepage
    I think that google's massive (over 10000 units) server farm (all x86) proves that the high end cpu's aren't worth it. Multiple low end CPU's do the same (if not better) job of one high end CPU. I think Google proves this point.
    • Re:google (Score:5, Informative)

      by Anonymous Coward on Monday August 06, 2001 @03:29PM (#2163754)
      Not every program scales well to multiple processors, though. In some cases it makes sense to have a single blazing fast cpu.
    • Re:google (Score:4, Insightful)

      by JWhitlock (201845) <.John-Whitlock. .at. .ieee.org.> on Monday August 06, 2001 @03:40PM (#2163832)
      I think that google's massive (over 10000 units) server farm (all x86) proves that the high end cpu's aren't worth it. Multiple low end CPU's do the same (if not better) job of one high end CPU. I think Google proves this point.

      How many fps does Google get?

      The article is in the context of buying a PC for personal use, and benchmarks using FPS, ray-tracing, kernel-compiles, etc. The idea is to pay attention to incremental performance (1.33 Mhz to 1.4 Mhz, .07 Mhz) versus incremental cost ($33? $100), and make sure it's worth it. Bottom line, buy cutting edge, get screwed on price.

  • Performance (Score:2, Redundant)

    by Phroggy (441)
    I've always been a large fan of using an army of small, low-powered boxes instead of one big expensive box. For one thing, if something breaks, everything else still works. For another thing, it's generally cheaper this way.
    • Re:Performance (Score:2, Interesting)

      by BigTimOBrien (203674)

      I agree.

      Well, one nitpick, it is generally cheaper in terms of hardware cost, but this option is more expensive in terms of operational costs. Also, the rent associated with rack space at a commercial provider can start to be prohibitive.

      So, three things start to add up here:

      • Operational Costs - People who can manage machines are expensive.
      • Rent - Renting space to house more smaller boxes may be prohibitive.
      • Network Traffic - More boxes means higher bandwidth requirements between these boxes.

      For some tasks, say a very large database with millions of transactions, it makes sense to pay the premium for even an extra iota of horsepower. For other task, such as web servers, it sometimes makes more sense to have many smaller machines. Also, if someone wants to start using this approach it usually pays to be able to autmatically configure a machine; otherwise, maintaining machines start to become unrealistic.

      • Re:Performance (Score:2, Insightful)

        by Deflatamouse! (132424)
        One addition to the operational costs: older machines (esp. 486s and early Pentiums) runs HOT! They are not as power efficient as some of the newer chips in terms of performance/power usage ratio. So if you take into account the amount of electricity sucked up by these little machines over its life span, you will probably want to trash your old systems and buy a new one (not P4s though!)
  • It depends on who's paying the bills. If you're an individual looking for a PC, even to play the hottest and newest games, you probably don't need and can't afford the newest processors. If you're a government or well-funded university lab, writing your own software, where the fastest results are critical, then you probably can't afford not to stay ahead of the processor curve.
  • One crucial point (Score:3, Interesting)

    by RainbowSix (105550) on Monday August 06, 2001 @03:31PM (#2163769) Homepage
    Sure, for most of us, save for games, a 166mhz processor is enough. I use that example because I run my laptop's AMD K6-2 333@166 (vcore 2.2@1.8, I/O 3.3@2.5) and it runs Enlightenment as well as I need, and at that usually at 0% load. For games, there isn't much of a gain from 1.33ghz to 1.4, as stated in the article. However, they don't make mention of people who NEED the full 1.4ghz. People who do rendering and other CPU intensive applications are the people who need to pay the premium. If you were rendring a scene or movie for money, the difference between the 66mhz and $25 could potentially be hours, days, or profits. Nobody buys a 66mhz faster CPU for $25 more thinking how much faster they can compile a kernel, but leading edge has its purposes.

    Of course, some people just like to brag, and ego can be worth $25
    • I'll ditto this. My wife runs a K5 (or is it a K6?) at 150 mHz. Runs all the web browser and Quicken she needs/wants.

      To answer your other point, why market the 1.4 to everyone if only a small market needs it? To encourage conspicous consumption (another in my series of references to Econ 101), the motivation of almost every American consumer.

      • To answer your other point, why market the 1.4 to everyone if only a small market needs it?

        Because those penile-enlargement products aren't working as well as advertised, so they must look for other ways to over-compensate.

  • They can be (Score:3, Informative)

    by chennes (263526) on Monday August 06, 2001 @03:31PM (#2163774) Homepage
    In computational fluid dynamics, where simulations run days or weeks, non-stop, maxing out the CPU the whole time, 5% faster is a lot. The price tag is actually relatively small when compared to the time savings that you can achieve. They certainly don't make sense for the average consumer, but for some simulations, they're worth every cent.
  • Hey, Win2k and Word2K aren't exactly flying but a few seconds wait is hardly intolerable.

    I think you can divide the world into two categories...people who play 3D games on their computers and people who don't. If you play mostly RTS games like I do (I still enjoy StarCraft) then I think you tend to fall to the bottom of the upgrade cycle.

    If you play mostly 3D games...it seems like you get sucked into ever increasing spiral of hardware needs. A new game comes out with a whole new bag of tricks (bump-mapped poly-textured fuzzy-logic nosehair) and you either need a good CPU to enable them or toss out your nVidia GollyGeeWhizForce and get whatever is the latest version.

    - JoeShmoe
    • I think you can divide the world into two categories...people who play 3D games on their computers and people who don't.

      You forgot engineers, video editors, etc. Speed does count there. However, I used to do the same sort of engineering jobs on a 386 -- just had to schedule my work so autorouting would run overnight, if I picked the right settings. 500 MHz will autoroute a similar board in 10 minutes, so scheduling is a little easier, and I can try more different settings to try to get the best routing. 1.4GHz might or might not get it done in 5 minutes (CPU cycles aren't the only limiting factor!). So if I was doing that everyday, a few hundred extra for the fastest system would improve productivity be cost effective. Two 500MHz boxes and a keyboard-monitor-mouse switch would improve productivity far more...

      Anyway, routing circuit boards is not my main job, and 500MHz is essentially instantaneous for everything else I do. But I remember when home computers were 8-bitters running games like the text-only D&D, and engineers got $50,000 Unix workstations, or else just a dumb terminal and a logon to the mainframe. Now the gameplayers need more power than most engineers...

      • I thought the question was primarily focused on home users. There are tons of business applications where speed = productivity. In which case, they are usually running something really high powered like a Sun/SGI or some kind of Beowulf thing.

        Now maybe there are people out there who do these kinds of things as a hobby, but then letting things run overnight isn't really a problem because you should have the same time demands with the hobby as you would with a job.

        I do a lot of MPG/Divx/MP2/etc encoding and I often have to let things run overnight (or in the background all day while I do other things). Again, not job critical so my slower processor is sufficient. If I had to make a living doing this and I was being paid per job (as opposed to per hour which would be QUITE profitable) then I definitely would pony up for a 1+ Gz system.

        - JoeShmoe
    • I only found RTS games after playing the free version of Red Alert a few months before the sequel came out. What the hell was I missing? Just as I was about to dump Windows entirely, I found this lovely game. (And then shortly after, found Baldur's Gate).

      Anyway, one of the reasons I was going to drop Windows games is that it would be much cheaper to get a console, and ignore the buy, upgrade, repeat cycle of 3d cards.

      Things worked well. Until I bought Black and White. My poor SLI Voodoo2 setup was no longer sufficient. (of course, the 600mHz proc. might have had something to do with that as well:) So, out went the Voodoo, in went a GeForce MX (or whatever the bottom of the line, cheap shit GeForce is) and it works. Quite well for B&W. But that won't be for long.

      Which is also okay. Turns out that I can use some of that ancient 3dfx crap in a Mame box. And if that doesn't work... Hell, for as much as the GeForce3 costs, I could at least one actual arcade game. Throw in a new motherboard, processor, etc. and I can get one of the pins that I've been wanting.

      And the people in the neighborhood (okay, the kids) much prefer the arcade games when they drop by. (No, I'm not a molester. I'm just immature. Arcade games are in the garage, and the door is open if any neighborhood kids are playing. And I have their parents' numbers.)

  • by supabeast! (84658) on Monday August 06, 2001 @03:33PM (#2163787)
    Unless a CPU is going to be used for high-end gaming with pure performance in mind, buying the high-end monsters is a waste. The money is much better spent on RAM (Especially in a Windows machine.), or faster hard disks.

    Even if the machine is to be used for gaming, the money is still better spent on a nice video card with a boatload of RAM, to compensate for the extreme sluggishness of a PC's system bus.
    • Not only gaming. There _ARE_ other CPU bound tasks. True, not many users out there are doing them, but they are out there.

      Someone brought up computational fluid dynamics - I'll bring up large statistical analysis problems, where large mainframes take a week, never mind many days to do a subset of the problem on a PC


    • Even with games its not worth it.

      Most games are targetted to a certain level of machine expected to be the norm when released.

      The higher the requirements the smaller the size of buyers. This is bad in the eyes of publishers.

      Internet connection and video cards are more important than cpu in almost any game.
      • True, but buying a CPU in the top end group (not necessarily the fastest, but close) does save some money in the long run since it gives your machine a certain amount of longevity.

        So buy a monster now and still be able to play games in 3 years.

        Pete
        • The Internet's sole purpose is to get porn and bomb making plans into the hands of children

          Porn making plans? "Insert slot A into tab B..."
        • So buy a monster now and still be able to play games in 3 years.

          Wow.. now there's an exaggeration if I ever heard one. Let's be honest: three years ago I dropped $1500 on a P2-400 with a TNT2 vanilla and an Aureal 4-channel soundcard. If you think that setup can play today's games, you are either out of your mind or haven't installed Max Payne :) I'm lucky to get 1.5 years outta my box before the games start looking like slide shows. I'm not really sure what the original poster was talking about anyways: games are the only category of software that actually push the performance envelope, and no, games aren't targeted at the current middle-end machine. They will run acceptably on such, but true gamers want all the bells and whistles, and for that you need the absolute high end. Think back to when Quake 3 was released: could the "middle of the road" do 32bpp on High Quality at 1600x1200. I think not; only some really powerful shit could. True gamers are hardly satisfied with their hardware 6 months down the road, to say nothing about years. So your partially right, you gotta buy that monster CPU. But it ain't gonna last you 3 years.

  • ... nice article. I like the $5.50/sec breakdown on the kernel complie. I guess this also relates back to the "its not the size that matters" argument. Sure sure 1.33GHz is probably a better buy. Usually your friends will go "whoa that's fast" with a 1.4GHz instead of "damn boy, you saved $33???? You're the man!" with a 1.33GHz =)

    Of course if you buy intel they should say "ahaha you are stupid." The advantage of this scenerio, however, is that they know you broke the bank with an intel proc so they won't hit you up for a 20 spot.
  • by IvyMike (178408) on Monday August 06, 2001 @03:34PM (#2163793)

    If your decision is between a 1.33Ghz athlon and and 1.4Ghz athlon, and the price difference is only $33, then of course it's worth it to get the 1.4Ghz! Otherwise, every time your friends use the system and say, "Wow, that's really fast! What is it, a 1.4Ghz?" you have to bow your head in shame and say, "No...it's a 1.33Ghz." You might as well throw Windows ME on it! When you're getting the hot rod of systems, it's not about bean-counting, it's about style.

  • Well Duh! (Score:5, Insightful)

    by gmhowell (26755) <gmhowell@gmail.com> on Monday August 06, 2001 @03:35PM (#2163803) Homepage Journal
    What hard hitting journalism. An amazing display of analytical prowess. I've had better stories rejected.

    Of course the top of the line stuff is too expensive. What the hell is there even to discuss with this article?

    (At home, I have a Celeron 466 or so on my Linux box. a PIII 600 or so on my 'doze box for games. Big frickin' deal, right? For the price of a processor upgrade, I can be running 1GB of ram in both systems. Through in another 100 bucks, and I've got more disk space than on the file server here at work (which is no slouch for what we do)).

    Guess what? Processors don't really matter anymore. Neither does any of that hardware. What in the hell is anybody doing with computers that requires all of this horsepower? Yeah, something will come out. But what, and from whom? Don't we have enough cycles to have incredible voice interfaces? No, because everybody (and by that, I mean Joe Six Pack, aka, my mom) needs M$ bloatware to do anything. It's because Quicken wants to do so much that it takes many megs of RAM to load. Why???

    Slashdot latest headline:

    Top of the line stuff gives marginal improvements for mega price increase.

    Christ, we knew that back when it was a 486-20 mHz vs a 486-25 mHz (and probably earlier). Christ on a crutch, how is this news?

    I think I know how stories are picked: each one is printed out. One of the editors grabs a stack and wipes. Whatever story isn't covered in it gets posted.

    Excuse me, I must go beat my head against the wall.

    (And please, anybody who wants to mod this down, I would much prefer it if you answer my question: why the fuck does this matter?)

    • You're absolutely on target about the situation. I recently built a mondo system, and found out a few things I had only suspected:
      The difference between a 750Mhz Duron and a 1.2G Athlon isnt perceptable in most programs, even Q3Arena.(With the same hardware)
      RAM makes a difference to 256MB, above 512MB you can get problems.
      Hard drive speed is now the major bottleneck.
      Try a KT7A-RAID (Cheap,now) with an array of 7200RPM fast drives, and you will be amazed at the boost! Be careful about shutting down, as those heavily cached harddrives don't get to write their info when you shut down! And the windows fix for that problem doesn't work.
    • Going by THAT Mac and Linux comparison then... I would guess that THEY would say to me.... Amiga? Why do you need a GUI that does multiple Shells (Command Line Prompts, with history), Multitasking, can look like you want it, can switch apps from 640 to 1600 screens without re-booting, can do a Disk-scan without stopping you from doing your work, can navigate through the directory levels faster than a CPU 10x faster......... Oh! I forgot, I'm accessing the Internet with a DEAD COMPUTER!! Drats, JK
    • (And please, anybody who wants to mod this down, I would much prefer it if you answer my question: why the fuck does this matter?)
      Those who do not learn from history* are doomed to repeat it.

      * history = The marginal cost of a 386DX-33 over a 386DX-25.
    • Re:Well Duh! (Score:4, Interesting)

      by cdlu (65838) on Monday August 06, 2001 @03:58PM (#2163959) Homepage
      Well, yeah. For people who actually do their research, this article doesn't much matter.

      The point is, though, a lot of people simply don't. A lot of my housemates, for example, have been having an informal rivalry of who can get the fastest system, and one of my housemates decided he'd win in a hurry by buying two >1GHz systems and 2 19" monitors at a cost of well over CAD4000 (about US$2600). Was it a smart move? No. He claims he needs the faster computers for his genetic algorithm work, but the 450MHz system he had before did the job fine. It still takes most of the night for his programs to run, the only difference now is he's a long way from waking up when they're finished instead of just about to wake up.

      Think about it this way, if this article didn't need writing, the hardware companies would not get away with the high prices they charge for their newest goods because everyone would be smart enough to see through the thin veil of little blue men dancing around a giant '4'.

      For the record, I am typing this on a 233MHz P-MMX which does everything I need it to do and then some, and continues to thrive as my primary system, allowing my money to go to more important things like eating lobster. :)
      • >>Well, yeah. For people who actually do their research, this article doesn't much matter.

        Good point. But is the slashdot audience the type for whom this is news?

        Nope. Says it on the masthead "news for nerds".

        (But yes, for the average manager/consumer, it is news. But why post on /.)

  • Most systems with high-end CPUs have the real bottleneck somewhere else (memory, motherboard, graphics). A lot of systems out there would benefit more from another 128 MB RAM than another 0.2 GHz of CPU speed.
  • I have an Athlon 800 (had it for quite a while now), play all the newest high-end games, and I have to tell you, for an industry that's about to pump out 2GHz processors, I haven't found even the slightest need to upgrade.

    If I was to buy a new machine now, I wouldn't touch anything above 1GHz... I'd go, preferably, for slower with multiprocessors....

    Even games, which are always bleeding edge (although that ruins the gameplay, but I digress) aren't running with the top processors. I say buy what you can afford - 1 level.
    • "Buy what you can afford"--that's what I'm going to do this fall, when I upgrade from a BH-6/overclocked 300A combo to something a little faster--an NForce motherboard and a zippy new Athlon.

      Of course, when what I can afford is in the vicinity of an Athlon 1.2...well, I'd be a fool not to take that. The way I figure, 1.2 gigahertz will run any game from now until they come up with holographic projectors for virtual reality. Or perhaps even now until the singularity, whichever comes first. :)

  • Once upon a time, the CPU was a hell of a lot more expensive than it is today. Before Intel had competition (and for a little while after AMD joined the party), their highest-end chip would cost about $800 and change, the next one down would cost around $600, and then the prices would drop off quickly. Back then, it made a lot of sense to buy a chip a rev or two behind the top of the line - the performance wasn't much different and you saved huge bucks.

    Nowadays it really doesn't matter that much. Intel chips are still more expensive, but nowhere near what they used to be, and there's only a tiny difference between the top-end Athlon and the next one down - and even the fastest AMD chip is less than $200. It's just so cheap now as to not make a difference anymore on the desktop.

    Intel still gets a premium for the Xeon processors, since AMD isn't really competing fully in the MP apace yet, but those will fall, too, over the next year or so as AMD competes in the server market.

    So if I'm building a system today, I'd buy the top-end AMD processor and build a nice system around it. But by the time all the parts arrive, there'll be a newer, faster, and cheaper processor out anyway, and I'll just have to cry over it. Such is the way of Moore's Law.
  • About the only use for high-end processors on Intel boxes (other than games) are high-end databases (like Oracle). And, consqeuently, most people would never run something like Oracle on an Intel box. It's usually on an HP box (HP-UX) or Sun. Ever web servers (even heavily hit ones) are NOT CPU intensive at all. I've got 20,000 pageviews a day going just fine a PII 233. It's the databases that are fat, and nobody in their right mind would run an enterprise-class database on a PC that you can buy at Wal-Mart.
  • by r_j_prahad (309298) <r_j_prahad&hotmail,com> on Monday August 06, 2001 @03:50PM (#2163894)
    Almost as quickly as Intel or AMD can release a faster CPU, Microsoft releases an OS that runs like shit on anything less. If the CPU designers don't keep pounding relentlessly away at Moore's law, we could theoritically have an OS from Redmond that won't run on anything.

    Wouldn't that be nice.

  • by isa-kuruption (317695) <kuruption&kuruption,net> on Monday August 06, 2001 @03:56PM (#2163946) Homepage
    You think this doesn't happen in other industries as well? For instance, you think a Lexus ES 300 is any better than a Camry XLE as far as performance? Okay the ES300 is 210hp and the Camry is 194hp... that's a 8% increase in performance yet it has a 20% markup for your wood trim and extra 2 choices in exterior colors! The same can be said about ANY higher end car compared to the lower end model.

    This is what happens when you have a capitalist government. The thing is, the companies know they can get a high price for the latest and greatest because there will be a certain percentage of us who will pay that price. Then, when prices "slump" a little, they will release a new chip that's faster and lower the price of the other chip. So now the "general public" gets those older processors at cheaper prices and that same group of gurus/morons will go out and buy the newest and greatest again. And the cycle of life continues...

    One reason a company makes the premium product higher is because they need to recover R&D on that product, however I don't see why this is in the chip market. I honestly feel that Intel and AMD "milk" the market for these gurus/morons knowing they will always buy the greatest. So they release a 1.0ghz and these people get that, then they release 1.1ghz and they get this one, etc etc. Although AMD has the 1.0 and 1.1 developed at the same time, they strategically release the products to the general pulic to maximize their profits. Of course, again, this isn't anything new... but it's painfully obvious.

    • by YIAAL (129110) on Monday August 06, 2001 @04:25PM (#2164159) Homepage
      This is what happens when you have a capitalist government. A capitalist government? We've never seen one of those, that's for sure. This is what happens when you have a (more-or-less) capitalist economic system. Poor people get a lot for a bargain. Rich people pay a premium to feel special. The premium that the rich people pay help companies charge less for the lower-line products that poor people buy. Everybody's happy. Rich people aren't forced to contribute (except by their egos) and no political bureaucracy dissipates the money being redistributed because it makes economic sense for companies to act this way. Hurray for capitalism! Thanks for pointing this out!
      • "The premium that the rich people pay help companies charge less for the lower-line products that poor people buy."

        Really?

        So it's got nothing to do with the fact that people who buy less-expensive cars are often looking for good value-for-money and competition (remember that ?) amongst car manufacturers mean they have to keep prices low if they want to sell cars.

        So why do companies sell less-expensive cars at all if they don't make money out of it ? Please, don't make out that by driving a Merc you're doing poorer people a favour.

        • It's called price discrimination. You segment the market to charge what the market will bear, but in different sectors. In tech areas, you charge early-adopter types a premium. That helps you pay back R&D faster. Then, as technology ages a bit, you go for high quantity and lower margins, while rolling out the still newer and fancier stuff to grab the rich guys all over again. Does it work? Hell, yeah. I'm writing this on a contemptibly cheap machine that would have been a supercomputer a decade ago. I do wish I drove a Merc, though -- so long as you mean a Mercedes and not a Mercury, which is unfortunately close to what I do drive.
    • True regarding the performance of the two cars. But if you drive a Lexus people think you're successful. If you drive a Toyota you're just another schmuck. The problem with your arguement is that you are disregarding how the two separate brands are percieved. Automakers do this with countless models. Honda/Acura Accord/TL, Nissan/Infiniti Pattfinder/QX4, Toyota/Lexus Land Cruiser/RX430. If you're worried about bang for the buck performance on a car, go out and get a camaro (ick).

      Anyway, your car comparison doesn't fly when put in the same context of the article. If you buy a 1.4 Athlon vs a 1.33 Athlon you'll end up saving about 6 second of kernal compile time (read the article). If you buy a Lexus instead of a Toyota you're buying status.

      Pete
      • I think he covered it pretty clearly. You're saying that it's a status-consciousness thing to drive a lexus over a toyota. He's saying the same thing, except that he's calling the status-conscious people morons.

        I'd also like to point out that you're not buying status, you're buying perceived status. Your personal echelon of society may be wowed, but there are those above and below you (not you personally, the hypothectical you) who will be completely unimpressed or (as the ads will have you believe) completely wowed.

        Personally, I can't imagine being wowed by a lexus, anymore than I would be unimpressed with a toyota.
        • True. Most of what I said is what the marketers want you to believe, but you have to take into account how status oriented America is. The Lexus & Infiniti nameplates don't even exist in Japan. They were created so the Japanese car makers could break into the market with their high end offerings. Like it or not no one would pay $70M for a Toyota LS430. A Lexus LS430 however...

          My main point was that percieved brand status doesn't translate to computer stuff (for the most part, there are still brand zealots out there), therfore his entire arguement is invalid.

          Pete
    • One reason a company makes the premium product higher is because they need to recover R&D on that product, however I don't see why this is in the chip market. I honestly feel that Intel and AMD "milk" the market for these gurus/morons knowing they will always buy the greatest. So they release a 1.0ghz and these people get that, then they release 1.1ghz and they get this one, etc etc. Although AMD has the 1.0 and 1.1 developed at the same time, they strategically release the products to the general pulic to maximize their profits.

      It's quite possible that they _need_ to do this continuously to recover the R&D costs and other overhead costs.

      Chip cad/simulation tools cost $0.5M-$1M per _seat_. Prototyping runs cost $300k+ each. The test gear and special-purpose simulation rigs aren't cheap either. A fabrication plant costs over $1B to build, which must be amortised over all chips manufactured there until the next major fab overhaul (typically only 2-3 years away).

      Chip design and manufacture is _expensive_. I seriously doubt that they're gouging as badly as you seem to feel they are.

      (In case you're wondering how smaller low-end x86-makers survive - they outsource their fabrication, which saves on amortized fab costs and capital at the cost of not having a fab process optimized for their chips (taking a performance hit). If they're wise, they'd also design their chips to either be more robust, or more easily tested, or both, to cut down on simulation/testing costs at the expense of performance.)
      • Forgot another ongoing cost - making new revisions of a chip to fix bugs or reduce power consumption or better tune the chip to the fab process used. Chips need just as much ongoing maintenance and tinkering as software does, even if the first version runs adequately.

        Again, just like software debugging, this sucks up a vast amount of time and effort and resources for the manufacturer.
    • This is what happens when you have a capitalist government.

      Welllll, I don't think it's an evil cabal of government officials and captains of industry hoodwinking the poor consumer. While this does happen in other cases, there's a much simpler explanation for this.

      Costs do not scale linearly. At the low end the fixed costs are dominant, which is why you don't see 2GB hard disks anymore -- it's just as cheap to make them 10GB or more. Conversely at the high end of cost or performance the marginal cost of just a bit more performance or capacity is great.

      Like the guy who mated the helicopter jet engine to a motorcycle frame. The amazing thing is that it isn't that much faster than a high end conventional motorcycle, something like 210mph vs. 190s. But building a bike that will go 210 vs. 190 is a much bigger leap than going from 90 to 110, even if it is smaller in relative terms.

      I expect that if you're aiming to produce a 1.4Ghz processor, a lot fewer will check out OK at the rated speed than if you are aiming for a 1.33 GHz processor, all other things being equal.

  • There is another reason to buy the high-end CPU that I haven't seen listed. If you are going to own the computer for 3+ years you'll get more milage out of that faster CPU....typing this on a three year old 233.
  • Oh great (Score:2, Funny)

    by Mr. Sketch (111112)
    Newsforge about something you may want to argue about with your boss or significant other.

    Like I need anything else [slashdot.org] to arge with my significant other about. We fight enough about [redhat.com] other [slackware.com] things [goatse.cx], but she just doesn't understand. Oh well.
  • Bureaucracy (Score:3, Insightful)

    by Xenu (21845) on Monday August 06, 2001 @04:09PM (#2164046)
    If you are working for a government or large corporation, you may be better off getting the expensive, cutting edge machine if you are going to be stuck with it for the next 5 years.

    Typing this on a blazing fast P5-233, and this is the _fast_ machine in my office.

    • P5?? So is that 233 Ghz? :)
    • Re:Bureaucracy (Score:2, Interesting)

      by spyderbyte23 (96108)
      My office's round of upgrades resulted in us getting 1 Ghz PIIIs with Geforce2 256 video cards.(We're a helpdesk at a university.)

      My boss pointed out that she didn't know when the next time she was going to get to buy machines was, and so she figured she'd try and fight obsolesence as long as possible.

  • Obvious answer: (Score:5, Informative)

    by proxima (165692) on Monday August 06, 2001 @04:16PM (#2164098)
    Are High-End CPUs Worth The Money?

    No.

    Now, for a better question. Are high-end motherboards worth the money?

    Every penny.

    In the many, many computers I have built and fixed (I don't know how many hundred..I never counted), one thing became crystal clear: don't skimp out with a cheap motherboard in order to buy that next higher-up processor.

    Motherboards are not created equal, not even close. In fact, from my experience, they are either the cause of good reliability or they are to blame for crashes and instability (in terms of hardware). Buying a good chipset put together by a good hardware manufacturer (Abit, Asus, etc.) is key to building a reliable system that will last several years of hard use.

    A good review site for motherboards will describe not only the features it has but how those features are laid out. A well designed motherboard has shorter interconnects and well placed components. Also, a motherboard should have a nice array of capacitors that keep maintain the electricity going to the processor. There should be ample room around the processor to stick the larger and better cooling cpu fans (another things never to skimp on). A heatsink and fan on some of the chipsets helps to improve reliability.

    But from my experience the best part about going with a better name is a reduced likelihood of getting a dud. I ordered a cheap Soyo motherboard to fit a K6-2 450 Mhz processor I had sitting around - I wanted a cheap computer. The first one was a dud, the second one was a dud. I ended up going with a different manufacturer and getting a 750 Mhz Duron. I had previously purchased an Abit with a Duron 700 Mhz and had no problems whatsoever. You pay about $20-$30 more for the motherboard, but it's definately worth every penny.

    In short, don't bother spending that extra $30 to get however many more Mhz, or even to get the difference between a PIII and Celeron or Athlon and Duron. More important than speed in most systems in reliability, and for that you should plunk your spare dollars into the motherboard and a decent heat sink/fan.

  • by rgmoore (133276) <glandauer@charter.net> on Monday August 06, 2001 @04:18PM (#2164109) Homepage

    It's rather silly in a case like this to look at just the price of the processor, disk, etc. You have to look at the price of the whole system and decide what kind of tradeoffs you need to make. Is $33 worth it for a 5% increase in processor speed? That depends on how much the whole system costs; if the system costs more than about $700 then the $33 is less than 5% of the system price and it may be worth it to pay more for the extra speed.

    The case when this really kicks in is with expensive proprietary software licenses. I've seen various programs that I might want to use in my work that have license fees in the thousands of dollars. In some cases that's the price per box, but in others there's actually a per-CPU license. If you're running somthing that costs $5000 per CPU, it makes sense to spend some fairly serious cash on getting the fastest possible processor.

  • ...on where the bottleneck is. You can't run a system faster than the slowest component, so if you're mostly swapping stuff on and off disk, your cash is better-spent on more RAM or a faster hard drive. You'll get much better performance gains for your money.

    Then, of course, there's always the option of simply overclocking your hardware. Put the motherboard in an insulated styrofoam case, flood it with mineral oil (making sure there's no air or water droplets remaining), and then hook up to a decent-sized compressor. You should be able to get the system cooled reliably to very low temperatures.

  • The reason for this pricing is to maximize the profit reaped from the "money is no object" buyer - the one who will say "I want the fastest chip you can put in a PC" and not worry about how much it costs. You'd be surprised how many of them there are, and how much of a chip manufacturer's profit comes from these buyers. These are the same people that spend $400 on a video card to get them 50% more frames per second than a $150 card. Again, you'd be surprised how many of them there are.
  • by Courageous (228506) on Monday August 06, 2001 @04:38PM (#2164244)
    I cost my employer a little more than 4 cents
    a second to employ. So, if a CPU costs 40
    dollars more, a mere 17 minutes saving to my
    time pays for the difference.

    C//
    • I'd rather see your employer spend the $40 more on a computer capable of posting HTML-formatted /. posts :)

    • Are you saying that your employer pays you 4 cents a second? Or are you a contractor hired out by your true employer to your customer?

      Regardless, 4 cents a second (times 3600 seconds per hour times 2000 work hours per year) adds up to about $288k per year. At that rate, I'd sure as hell hope your employer was maximizing every second of your productivity.

      That'd be a good justification for installing workstations in the bathrooms. Or maybe toilets at the cubicles.

      • That'd be a good justification for installing workstations in the bathrooms. Or maybe toilets at the cubicles.

        Hehe. Nice idea... But how well would it go with the "paperless office" ideas? About the first idea... umh... having work stations with Internet access in restrooms sounds kind of kinky.

  • Price?? (Score:2, Insightful)

    by crandall (472654)
    High price? What are these people talking about? My 1.4 ghz athlon 266bus cost me about 190$ USD. Compare that to same time last year, when the top end AMD processor was 400$, and the top end Intel processor was 600$. Processors expensive? Maybe if you live in a trailer park.
  • by Anonymous Coward
    One of the things I like to do is compress mpeg2 video down to mpeg1, both from a digital camcorder and/or dvd's I have to watch on a plane (my laptop is too old to have a dvd drive and doesn't have enough cpu to watch dvd's anyhow).

    Compressing 3 hours of video and adding subtitles on my dual p3-450 was 18-20 hours.

    On a 1.4ghz athlon it's about 6.5 hours.

    YOW...definately worth the upgrade!
  • My K6-3 337 + EDO ram which costs much more than a 6x faster Duron 800 w/DDR ram is still plenty fast for me.

    obviously I'm not running windows 2000.

    building a new system? always get a the low end of the curve CPU (currently a 1ghz athlon) and a high end motherboard plus lots of ram, you'll be much happier with everything but your dnetc bechmarks.

    if that cpu isn't long enough for you you can upgrade it later to a 1.5ghz model when those are old and cheap one year from now.
  • .. I'd say high end CPU's are certainly worth the money. The company isn't "wasting money" for people down time - I can get back to coding instead of waiting for the compiler.
  • by plopez (54068) on Monday August 06, 2001 @05:13PM (#2164489) Journal
    Most of my work these days is in databases. It is MUCH more cost effective o spend he extra money on the fastest memory, memory buses, disks and controllers you can get. Most Intel type cpus spend an inordinate amount of time waiting on IO. Faster IO means a faster system. In addition, on the low end Wintel systems, SMP is a joke. I have yet to see a system running with more than ~75% utilization per chip. In addition, in database systems, Oracle and MS both charge by number of processors and MHZ. So going single processor with a slower chip can save a considerable amount of licensing costs!

    For a DB system the rule is 'fast disks, fast memory, fast buses, fast controlers', for heavy network traffic (lots of web hits), get the fastest networking you can afford.

    And remember, MHZ is only part of the equation on processors. If you really need (and few people really do) a fast chip, good and large L1 cache is a bigger win than raw MHZ.

    My $.02
  • If the processor is the thing you wait for then you need the fastest processor you can.

    I personally wait for my internet connection most; secondly I wait for the system to boot up/shutdown (that's a lot more to do with disk throughput); third I wait for my graphics card (I've got an ancient Matrox G400); last of all I wait for my processor.

    And that's the order I would upgrade for performance, given the choice. But if I spent all day burning processor cycles doing some compute intensive operation, then I'd upgrade my processor.

    The RAM speed doesn't make a lot of difference, only a couple of percent. Not having enough RAM can have a much bigger effect- I recently added 256 meg- and the system is now noticeably faster- I was using up my RAM and the system was swapping stuff out. (RAMs at 20c/Megabyte it's a good time to buy. [Also, Quake III was able to use the extra RAM and gives a higher framerate ;-)]

  • If you use the machine for actual work which is CPU bound (simulations, rendering) just calculate what your time is worth and how much you will save with the faster model. if number of hours saved over the lifetime of the machine multiplied by the cost of those hours is greater than the price difference the go for the faster one.

    You will find that it will take a very high price difference, not to justify a faster processor in such cases.

    However most people do not run anything CPU bound so they should find a cheaper model.

    Where I work we generally buy the fastest (dual) CPU workstations we can get simply because it makes finacial sense. We constantly run simulations or calculation heavy software. It only takes a saving of 1 designer day over the lifetime of a machine to justify a $100 price increase.
  • Per CPU licenses (Score:2, Informative)

    by cDarwin (161053)
    If you're paying through the nose for per CPU licenses then it often makes sense to get the fastest processor your application runs on.
  • With the newest technology, you're the pilgrim taking the arrows to see what quality control may have let slip (early adopter syndrome). P60's that double as hotplates, Zip dirives with the "click of death" come to mind. If you buy the cheapest available (most often the oldest), you run the risk of technical relevance and quality of support (why is it so darn cheap again?). I like sega, but if you don't own a Dreamcast, do you want to sink $49 into one at Xmas "just 'cause" when that could be a Playstation 2 or GameCube game?

    I don't think Joe Average consumer goes wrong with any technology buying somewhere to either side (or on) the middle of the road. Taking the leading edge or the trailing edge is the sure way to get taken as a consumer.
  • I guess that is why I bought my 1.2 Ghz Athelon. It was less than the 1.3 at the time and way less than the 1.4. The thing to also look at is to make sure you get the 266Mhz fsb with the athelon instead of the 200. It makes a little bit of difference.

    I saved money on the cpu to spend on memory, and hard drive. Which is used more than raw cpu power. Fact is that most people could get by pretty well with 600Mhz cpus. Even most gamers would be fine with 600Mhz systems. Unless you are doing some serious heavy duty gaming, or super intense graphics or scientific number cruntching then 600 is plenty fast. (or some type of emulation)

  • .....that there is 6 times as many posts here on /. than on the original Newsforge article..... Does that mean /. is running the 1.4Ghz and Newsforge the 1.33?
  • The subject sortof sums up how I buy motherboard/processors/etc. for my home systems. When Intel (or whoever) brings out their latest chip, I find great deals on the previous generation chips. I buy those chips at the higher clock speeds and then, lately, I'll use them in an SMP configuration. By the time I feel a need to upgrade again my systems are about 4-5 years old. However, I might just break with tradition soon and jump into a dual-Athlon board. Even doing that, it looks like I'd still save money over the latest Intel offering.

    And since Windows only gets used for (infrequent) games, why would I want or need to have the latest and greatest Intel space heater sitting under my desk?

  • ... it's the best value. I bought dual P2-450's 18 mos ago. They were the best value I could then. They were already old CPUs. I'm still using the machine today with 384MB of RAM. I only wish I had more speed on a small number of occasions (I don't get to play Quake 3 as much as I would like!). I have a good graphics card, lots of memory and fast hard drives, and I rarely feel like I need one of these new fangled 1.5GHz CPUs. I'm thinking of upgrading, and it looks like P3-850's would be best fit for the motherboard. I have a Tyan Tiger 100 (1832DL), and I'm hoping that Tyan releases a BIOS update so that I can run the P3-1GHz 100MHz FSB... but they will need to drop another $100 before I go there (probably about the time P4-2GHz comes along).
  • YES! (Score:2, Insightful)

    by Bitmanhome (254112)
    Like the noisy AC said back there, ALL CPUs ARE CHEAP TODAY! If the computer is for work, then even the most expensive computer will easily pay for itself. If the computer is for home and doesn't earn you any money, then the only question is: How much do you wanna spend?

    It's not a question of price/performance, it's of price/happiness. If the dollars make you happier, then keep them; if the megahertz makes you happier, then buy them!

    -B

  • by leonbev (111395) on Monday August 06, 2001 @08:40PM (#2165439) Journal
    A long-standing guideline for purchasing CPU's has been to buy 1 notch below the absolute latest in technology. That way, you can get about 80% of the performance that the newest product offers, at about 60% of the cost. That way, you can get the best price-to-performance ratio, and have some money left over for other computer components. The "cutting edge" technology almost always has at least a 20% price premium attached to it, and should be avoided whenever possible. Save that money and spend it on something else, like more memory or storage.
  • It's a lot easier that way. If you know the right bars to look in, you can shop around and get a pretty good deal right now. Of course, if your SO comes with a lifelong service contract, this is a WAY toi expensive upgrade path.
  • Interesting article, but I have to take issue with the monitor analogy. First off, the math is wrong: Going from a 19.9-inch to 20.0-inch viewable area is 1% more screen, not the stated 0.5%. (Screen area scales as the square of the diagonal; a 20-inch screen is four times the size of a 10-inch screen, not two times.) Second, chances are good that these two monitors have differences besides the extra 0.1 inch of viewable diagonal. The more expensive monitor could well have higher refresh rates, better color-calibration options, or other features that drive up cost.
  • When you start working and paying far a car, paying your own rent, paying for health insurance $166 is nothing.

    Unless you believe in conspiracy theories of course.

The more cordial the buyer's secretary, the greater the odds that the competition already has the order.

Working...