Slashdot is powered by your submissions, so send in your scoop


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

Tom's Reviews Kryotech's 1000MHz PC 158

GenBradly writes " Tom's Hardware gives a review of Kryotech's newest model, the SuperG. They claim this new unit allows you to stablely overclock your Athlon to 1 GHz. Though the unit weighs 70 lbs, it's much quieter than previous [Kryotech] models, only slightly quieter than elevator music. It may be a bit expensive but I would do a lot for a 1 GHz computer." Lots of pictures, lots of specs, lots of tests.
This discussion has been archived. No new comments can be posted.

Tom's Reviews Kryotech's 1000MHz PC

Comments Filter:
  • by Anonymous Coward
    Alpha's are running in the 800+MHz range (at least, the newer 21264s are). But they aren't the fastest, MHz for MHz. In fact, the per-clock execution speed of alphas is quite poor for integer and moderate for floating-point. MHz for MHz, MIPS are the fastest normal processors (See the top 500 list to see some absolutely insane hitachi machines with only a few moderately-clocked processors which are probably the fastest in existence). But yes, Alphas do run at insane clock speeds and have for some time.
  • With all the hype to break the 1GHz limit, aren't we forgetting something? Those that are thinking about purchasing this monster, please think about your electricity bill, first.

    This is like having a second refrigerator (a small one, but a refrigerator nonetheless) in your apartment/house/cardboard box. Add to this the habit of most geeks (like me) keeping their computer on 25 hours a day, and you get the idea...

    I'm sure that we won't have to wait for 1GHz processors from both Intel and AMD for more than 6 months...


  • Yeah, I know. That was one of the biggest mistakes with the 386. Intel knew that FPU performance was important, and they were redesigning the FPU anyway (AFAIK, the 387 was incompatible with the 287). They should have abandoned the stack based FPU at that point.

    People always talk about how great Intel's FPU speeds are, but that's only relative to other x86 chips. And now they're not even top there.

  • Just a note: Kryotech bundles nothing with it's systems. It's all barebones. You get the case, cooling equipment, mobo, and cpu. That's all Jeremy
  • hahaha

  • Not letting an AC have a first post doesn't really sound like censorship....they can still say whatever they want. Not only that, they have the option of getting an account any time they like. It's even free.

    Booting a person for what they say...well...that is censorship.

    As for moderation...well...moderation on usenet is basically canceling posts. Is that censorship? Without that "censorship", alot of groups would be like the *.alt groups many of which are so full of spam they are practically worthless.

    I'll drop this issue here...but I'd have a hard time defending "First Post!" messages with a straight face. To each his own.
  • lest we not forget, up until fairly recently ALL computers cost $2500.

    Here is a short list of computers that I have purchased with $2500:

    8086 XT with two floppy drives and a green screen
    80286 AT with a hard drive (can't remember how big)
    80386 Compaq clone with 1 meg and a 65 meg hard drive
    80486 with 16 meg and a 400 meg hard drive
    pentium 90 with 64 meg and 1.2 gig drive
    dual pentium pro 2/512 meg and 2x4.5 gig scsi
    dual pentium III with 1 gig and 3x9 gig scsi

    so--$2500 will always buy you a "state of the art" computer....and the state of the art changes about 3 days after purchase/delivery. Mainly because it was state of the art when you were contemplating ordering it--by the time it is delivered there is always a faster/bigger/badder machine available.

    I don't advocate spending $2500 for one of these cooling units--just wait three to six months and you will get a much bigger bang for your buck....but then, you could always say "wait a while--the new xxx-IV is coming out next month"

    so, do what you like!

    just my $0.02 (collect 125,000 of these and you can buy a new machine!)
  • What about FibreChannel ???
  • AMD has said that they stand behind the Kryotech systems. Therefore, the parts are being run within the manufacturer's specifications, and are not "overclocked".

    If you do the same thing at home, it's overclocked.

  • Gosh, if machines keep getting faster, maybe we can reincarnate the old Symbolics machines, where everything including the operating system, is written in Lisp. It's just a bit hard on your TCP retries when the garbage collector fires up. :-)
    Yeah! And if they really get fast enough, we can write whe whole system in perl :-) Just think of it! A perl compiler writen in perl!
  • Random thought: is not Super G also an olympic skiing event?

    Might be hard to trademark anyway ;-)
  • Actually, hard drive speed have been increasing (thus the need for faster interfaces - U2W SCSI / ATA66) One important factor in hard drive speed is the spindle speed of the hard drive. I have a few 5400 and 7200 rpm hard drives. The 5400 rpm hard drives are kinda warm after a few hours of use, but the 7200 rpm drives gets hot after a few hours of use. Now a 10000 rpm (currently the fastest afaik) hard drives... most of them require fans to keep them cool or at least some decent air flow. If you try to go any faster... say... 12000 rpm or 14000 rpm, umm... well... I think the hard drive will probably need it's own refrigiration device. =)

    The importance of hard drive speed also depends on what you do. If you are running a webserver or maybe a database where you constantly need to access the data, then hard drive speed is very important. But if you are say... playing quake, then hard drive speeds are not that important because the hard drive is only used to load up the game into memory, once all the data is in memory, then the performance is mostly determined by the cpu speed and the video card.

    There is no statute of limitation on stupidity.
  • Why is he testing games with the G-Force. It may be the fastest thing out there but the drivers are not done. Expecially when it comes to 3dnow in the drivers. Tom did this in his last review of the coppermine. He tests his systems on graphics hardware that is not yest out nor is it complete. So far as I can tell the g-force isn't out and he shouldn't be using it to test other components. It seams ridiculous that an athalon can score much higher on the fpu mark but only hold its own on fpu performance in 3d games. The 3dgame benchmarks should be dismissed. They are invalid.
  • My review?
    It's 1,000 MHz.
    I think that's all I need to say.

    If you think you know what the hell is really going on you're probably full of shit.
  • These complaints of "only a 33% increase over dual 550 Celeron" and "what about the electric bill" seem a little picky to me. As to the former, an OC'd Celery has exactly no warranty, whereas the Kryotech system has full warranty even at 1000Mhz. And everyone knows that high-end purchases suffer from diminishing returns syndrome, but if you want the fastest you gots to pay the mostest. As for the electric bill thing, I wish that was the bottleneck in budgeting my money. It seems to me OC'ing is about testing the limits of manufacturing processes, discovering quirks of BIOS settings, etc. As for the HD being the true bottleneck, depends on the app; 256Mb PC100 with an Athlon ought to give a pretty decent throughput even for 3D processing apps. One more thing: when the Athlon 1000 comes out, you'll be able pop it into your Kryotech setup and run it at 1200 or 1300 or something, i.e. while expensive, it does offer some upgrade room.
  • by Anonymous Coward
    This post is still a cut above most others in this discussion. It at least doesn't pretend intelligence.
  • Considering that AMD is SCHEDULED to introduce a genuine 1 ghz processor in the first quarter of next year (jan?) by the time anyone buys the athalon cooled part, builds his computer around it, tweak it, it'll be early to middle december. Then a month later, the real thing comes out $1000 (at least) cheaper. How are you going to feel about that???

    Also, may we have a poll here about what we need a 1 ghz processor for besides playing quake and webhosting and cad?

    Please visit [] - You can donate Food and Medicine for FREE to Save Children. The donation is fully paid by corporate sponsors with the money they would have spent anyway on advertising. There is no charge to you.
  • All I can say is... Don't feed the troll!


  • There's quite a bit more to it than just HD access time -- though that certainly is a limiting factor. But it starts much "earlier" in the memory hierarchy, even with the caches, and to a lesser extent the DRAM. As clock speeds increase, the /relative/ time penalty of a cache miss (or a memory miss) also increases. Given that this overclocking scheme doesn't do anything to improve cache or memory hit rates, the end result is that the actual performance increase is much less than a comparison of MHz would suggest. Each instruction that does not find the necessary data on the first try will have to wait longer, relatively, and this damps performance gains. Patterson and Hennessey's _Computer_Organization_And_Design_ gives a good example of this (page 567 in the second edition). Doubling the clock speed of a computer, without changing its memory systems in any way, will record only a 41% performance gain. While this ignores some considerations, it's a fairly good approximation of the real world behavior. Anyway, I appreciate Bigger Faster Better More as much as the next guy -- and damn! 1 GHz just rolls off the tongue! -- but it's not going to be _that_ cool. Yet. But I still wouldn't mind trying one out :) John
  • >> does having a dual 550 make that much difference to a single 550

    it does, with some proper operating system. remember, win98 or similiar rubish won't work on dual processor machine. my personal favourite is linux, but u don't get a lot of games for it. (at least, not my kind :-). ) For me Win2000 works fine, but none of pre-releases had true 3D support, so if u are happy to run directx 3.x, suit yourself. in fact i think win2000 official release will have directx 7.0 with true SMP support... that should satisfy the every gamer, keeping in mind, that one can set up a game server and play at the same time on the same machine without serious losses in speed. in fact, u might not even notice it. playing net games is even much more fun, when, on win2000, one cpu is supporting the game, meanwhile other one works on your network status, os and other... good suggestion, try using some mean 3D video card (geforce, tnt2, voodoo3, etc. will do), to keep the load off the processors. i noticed with shit card (intel i740, with whatever memory on it)u have some great losses in speed, due to the fact that both processors are working to satisfy your card requests.

    so, i think, there's still much to go in SMP gaming, keeping in mind that the game itself should support SMP, not just your os alone...

    any questions, e-mail me at

    'cognito ergo sum', Aristotel
  • mcelrath wrote: Ask anyone -- they'll tell you alphas are expensive, but no one actually knows how much one costs

    Hm, could that be because I can't just stroll into my friendly neihborhood computer parts store and pick one up? Typically, when you see people rant about how cheap Alphas are, they're talking about last-last-year's chips. Let's face it: Alphas are simply not as generally available as x86 stuff. Please prove me wrong.
  • What do i multiply with ? 1.6 or 2 ? or is it lighter than 70 kg ? I mean if it weighs more than 100 kg then ouch ! I have moved a lot of UPS's in my day (they are heavy as lead...)...

    -"Don't mess with that guy over there !"
    -"Is he a martial expert or a body builder you mean ?"
    -"Worse, he is a geek that uses a Kryotech cooled computer as a notebook..."

  • Certainly, multimedia apps do benefit from caches. I may have given the wrong impression in my previous post (which was rather rushed because I was trying to finish it before I had to go out). If I did mislead, I apologize.

    As you have said, certainly the programs that process the data will benefit from the cache. This is true of almost any program one is likely to run, and multimedia apps are no different in that respect.

    However, that is less true with some multimedia data. Yes, the data does benefit from a cache. The thing with heavy multimedia use is that the amount of data you have to get to the CPU is so much greater than for most other apps. And the turnover of data is greater. Caches work best with the most localized code and data - ie the stuff that gets accessed repeatedly. Caches are effective because if you access something 100,000 times, and 99,999 of those accesses are from the fast cache and only the first 1 is from the slow main memory, the *average* memory access latency as seen by the program is the cache latency, not main memory latency. Thing is that the multimedia data tends to stay around for a shorter period of time, so the main memory access has a greater effect on the average latency. So maybe the average comes out to 2 or 3 times cache latency instead (still *far* better than no cache, but less effective than the more localized non-media data). Also, the greater volume of data you have to get from memory may mean that the memory just can't supply the data fast enough to the CPU. Improving the bandwidth of the memory can help this (this is the approach Rambus takes), or lowering the memory latency (a la PC133). That's why Intel moved the PII family to a 100MHz bus a while back - the 66MHz wasn't able to supply data to the CPU fast enough. And similarly with the Coppermines moving to 133MHz.

    I'm not saying (and didn't say) that the rest of the computer ought to be as fast as the CPU speed. The technical reasons you mentioned are very much going to prevent that. And besides, there's probably a diminishing return as you approach equality between CPU and memory speeds. What I am saying is that in the end you have to try to keep everything balanced, so that no one part becomes too large of a bottleneck. Which means improving I/O speeds as well as CPU speeds.

    And just FYI, I'm not into the conspiracy theory thing either :-)
  • I believe the conversion is 2.2 lbs == 1 kg
  • Actually, there are 64 bit 66MHz pci-busses in use already in high-end computers.

    In theory that's 4x the speed of a "PC" pci bus, but since the overhead increases with higher speeds you may get 3 - 3,5 x.

    Now, why doesn't Intel or some other chipset-company implement *this* in standard "PC" hardware? =)

  • Seeing as intel and AMD ar both working on 1 ghz chips, I wonder how many people actually buy these...

    Also, wonder how fast these things will be able to run a chip build to run at these currently insane speeds? :)

    bash: ispell: command not found
  • The only reason that PCI was clocked at /2 FSB was that it used to be 66MHz and most entrylevel hardware are made for the 32bit 33MHz PCI standard.

    Since that still is the case the PCI bus is still clocked at 33 MHz if you run 66, 100 or 133 MHz FBS.

    If you run at non stadard speeds, like 75, 83, 112, etc, you *overclock* the PCI bus and not all PCI cards can handle that...

  • Exactly right. Basically, the Alpha design philosophy is to get their performance from the clock rate. So they partition the work of the CPU into tiny pipeline stages, which allows really high clock rates. On the other extreme, you get the PPC designers. They decide to do a lot of work each clock cycle. The result is that the amount of logic in each of their pipe stages dictates a much slower clock rate. Modern x86 chips tend to be somewhere in-between in philosophy.

    Neither design philosophy is necessarily better than the other, performance-wise. Though going for MHz is better from a marketing standpoint :-)

  • Larger HD's lead to faster speeds too.
    If you double the density of data on the disc, you automaticaly double the amount of data that travels under the head in a single revolution.
    That leads to a doubled transferrate without increasing the speed of the disc.
    You can't simply increase disc-speed, since that leads to higher temperature, shorter life and higher noise.

    In a server, the temperature and noise isn't critical, since you usually got effective cooling and a soundproofed serverroom.
    But in most cases, it is smarter to use a real RAID system for speed.
    That way you'll also get redundancy when your disks break.

  • by Glenn R-P ( 83561 ) <> on Sunday November 21, 1999 @06:47AM (#1515169) Journal
    | Seeing as intel and AMD ar both working on
    | 1 ghz chips, I wonder how many people actually buy these...

    Why not? They could use the Kryotech to refrigerate the 1GHz chip and overclock *that*,
    couldn't they?

  • Seriously. I've got a 300 MHz AMD, and the CPU speed suits me just fine. Meanwhile, my multiple 7200 RPM SCSI drives, NVidia TNT video card, and six expansion cards are all generating heat like crazy. I've got a temperature probe on the video processor, and it hovers around 110 degrees Fahrenheit with the window open. I've also got five fans in this beast, so it isn't lacking in air flow.

    I want a Kryotech system that cools the whole case. Not chilling the CPU to sub-zero, but cooling the entire case to 50 degrees or so. I would be willing to pay serious money for such a system. Lower overall temperature prolongs life, reduces failures and errors, and can even improve performance.
  • Can I get one of those Suns for the same price as my "peecee"? Jackass. Quit trolling, if we could afford "real computers" do you think there would be any reason for using "peecees?" PC computing is all about getting as much performance as you can from what you have. Second, your "peecee" thing gives you away, even though you are an AC.
  • because i could pop that real 1GHz CPU in and run it at 1300Ghz or something.
  • Apple troll. (Insanely great?) The PPC FPU arch is crappy in the 603/G3 series, and still not as fast as Alpha in the G4 series. G4 has altivec, but Alpha has VIS (I think.)
  • OK, are you people done raving and foaming at the mouth? Deleting posts, banning users for just first posting, preventing first posts.. Does the term CENSORSHIP mean anything to you?

    Moderation is not censorship, just filtering. Browse at 1 or 2, and you'll never see a first post. Or an AC post for that matter, unless it gets scored up on it's own merit.

    First posts are a neccesary evil. You don't have to read them. You can block them WITHOUT censoring. So lets not get all huffy over it!
  • Read this [] please. I wrote it, on how censorship is NOT the way to deal with first posts. Karma will bite users for it, you can filter the AC's.. So where's to be gained by censoring?
  • The real bottleneck is in the keyboard/mouse!
  • Ace asked:

    Also, may we have a poll here about what we need a 1 ghz processor for besides playing quake and webhosting and cad?

    Sound and video editing. That's what I bought my Cool Athlon 900 for.
  • I get real tired of the few times I actually read the comments being full of flamebait and trolls. I'm sick of Slashdot being a continual form of immaturity battle...

    Welcome to the reality of a public forum. Terrifying, isn't it?


  • if a slow processor can be overclocked to 1 GHz, by supercooling, then upgrading the processor to something that naturally does 1 GHz, and THEN applying the same cooling, should give you an AWESOME speed!

    is a 700MHz processor (the one that was frozen) really that slow? It looks like freezing is only ever going to get you something you are going to get in a few months anyway.

    big deal.

  • Does that mean I'd have to let my wife win and get rid of my 3640 sucking down 30 amps from the special twistlock plug we had wired?

    The Open Genera Alpha implementation listed in the third reply is available. There is some talk of getting things moving again. I for one would like to see it come back to the land of the living. It's been the best rapid prototyping environment I've ever worked in.
  • Kryotech does sell a case just for that. I can't remember what they say it'll drop the temp to in your case, but I do know it's no where near the -40 (due to condenstation).
  • In order to overclock an Athlon, you actually have to take apart the container the chip is in and (de)solder connections. So, for your average company that would overclock chips in the first place to charge extra for a faster processor, I really doubt that the cost of overclocking it is worth the extra money they'd get out of it.
  • Geez, most trolls at least try.
  • It is most amazing to me how shocked we are all at the idea of a $2500 computer. I remember when I bought by 'badass' P120 with 32 rocking megs of RAM, a 1.7 gig HD and a shitty 17" monitor that blew out 15 days after purchase for $2600. All that was one of the cheapest ones out of the old favorite Computer Shopper. Now we are amazed at a similar price. Kinda weird.
  • About a year ago I replaced the motherboard in my server with a dual PII-400 board. The board+CPUs+RAM ran me just a little bit more than a thousand (I kept everything else - disk, monitor, etc...)

    That was a year ago. Although I haven't checked the prices lately, I would expect a dual PIII-500 to run for about the same. According to this article, the complete system runs about $2500, so I'd be surprised if a complete dual PIII-500 server can't be had for the same amount of money.

    Well, I got curious, and looked up VA Linux's prices. They quote $3600 for a dual PIII-500, but that's with a separate SCSI controller, and an 18 gig SCSI disk. That's a thousand bucks right there. I wasn't able to find what Kryotech puts in this machine, I'd be very surprised if they bundle something similar. Also, VA Linux usually comes out a little bit on a pricey side, and almost always you can get a better deal on the individual components elsewhere.

    If you're looking for a cheapest way to boost the power of your existing system, it'll definitely be cheaper just getting a motherboard+CPU+RAM, and reusing the rest of your equipment, instead of buying a complete system like that. You'll probably get better performance too in a multi-tasking environment. A single 1 GHZ CPU would will come out on top only in situations that involve linear processing. Try benchmarking 'make -j 4', for example. A single CPU won't buy you much there.

  • You are assuming a new bus is necessary. PCI actually works rather well. It's 32 bit wide and 33Mhz, which is over 1 Gbit/s -- enough to saturate a gigabit ethernet -- not that you'd want a gigabit ethernet on a standard PCI bus. There are 64 bit wide / 66 MHz PCI buses as well. You don't really need them on a home computer or even a workstation though.
    Oh, and btw, ISA does a very good job as well. For what it's being used (sound cards, modems, etc.) you don't need high speed.
  • It really doesn't seem to me that such drastic measures are neccessary, it only takes about half an hour(if that) for the dumb stuff to get sent down to -1. If you are browsing at -1 you are just asking for the muck. I think the current system is great.
  • Not true. mhz is irrelevant to the gamer-d00dz!@#

    You see... its FPS and ping time that make the man, baby. The game the FPS and ping are calculated in changes from time to time. Currently it is Q3 Demo Test.

    Of course, this is even more ridiculous than bragging about chip mhz speeds, especially when you're talking about the difference between 95 and 100 fps.

  • Gosh, if machines keep getting faster, maybe we can reincarnate the old Symbolics machines, where everything including the operating system, is written in Lisp. It's just a bit hard on your TCP retries when the garbage collector fires up. :-)

    Hey, wait a second. Isn't this what the Java folks keep waiting for? :-)/2

  • yeah dude, doom IS the best thing to happen to us since intellivision
  • Let's face it: Alphas are simply not as generally available as x86 stuff. Please prove me wrong.

    Agreed. Sad state of affairs that one vendor (Intel) dominates so completely. I dream of the day I can walk into a computer store and see Alphas next to PPC's next to AMD's next to MIPS', all running the same apps on Linux, and be able to evaluate them for myself.


  • Just grabbed the retail package cheap at a local store and tried them out at 550, succesfully with a bit of cooling.

  • I believe the conversion is 2.2 lbs == 1 kg

    1 lb == 0.45359237 kg, to be anal ret... exact.

    So that means, 70 lbs == 31.7514 kg.

    (Yes, I know the 70 lbs figure is an approximation to begin with, so we don't give a f* about the grams... But I'm a physics graduate with a prejudice against approximations. :) )

    "The wages of sin is death but so is the salary of virtue, and at least the evil get to go home early on Fridays."

  • Actually, according to the Bogomips mini-howto's bogo guestimation chart [], the K6-2 should be more like 21875x faster((350*2)/(8*0.004)), at least as far as bogos go. So even assuming bogos are wildly innacurate, the figure your benchmarker gave is pretty dang close, or at the very least an underestimate of the difference between the two chips.
  • Why is this of any relevance to anyone who isn't runing a non-multithreaded OS (ie Win9x)?

    Would we not be better served with dual- and quad-processor Athlon motherboards? After all, a 1GHz processor is going to spend a lot of its time waiting around for the RAM to deliver data.

  • While there is some value to a simple objective test of how fast Quake3 will run on the best system that can be put together, what Tom offers is comparisons between different processors. If these comparisons largely depend on the graphics card (and possibly its drivers) instead of the processor, then you are not getting the whole picture.

    We cannot reason ourselves out of our basic irrationality. All we can do is learn the art of being irrational in a reasonable way.
  • Mulitmedia does benifeit a LOT from caches. Often you must process (decrypt, convert) multimedia data in the processor, so the blocks of data being worked on are in the L1/L2 cache while the processor works on the data.

    Also, the programs that convert this multimedia data also live in the L1 cache. Since mulitmedia programs are about moving and processing large amounts of data in a loop, it does well in an L1 cache.

    Also, saying the rest of the computer should be as fast as the processor is wishing against hope. Lots of really nasty issues occur (feedback, crosstalk) when you try to drive wires on a motherboard at the speed of the CPU. The cost of designing and creating motherboards that can overcome this limitation are REALLY expensive and difficult to design.

    So no, the way the computer is designed is not a conspiracy to rob us of computing power! It actually makes computers affordable and faster than they otherwise would be.
  • Any moron that would spend their money on a box that contains basically a refrigerator oughtta have their head examined.

    I'm sorry, but there's no need to go that extreme for a couple of MHz. Sure, it's cool and all. (no pun intended) In a few months, CPUs will run at that speed without all that crap.

    PCs are supposed to be getting greener as much as they are faster. Somehow, I don't think this is in following with that philosophy. Besides, as much as it cools the inside, it *heats* the outside!

    It's cute, but I wouldn't want one.

  • How can the bottle neck be in CPU performance....It is a 1Ghz processor, How fast is the video card processor running, 400Mhz tops. Think about what you are saying before you type it, that way, you may get some facts right.
  • There is no "one true bottleneck" that limits computer speed. Some apps are limited by disk I/O. Some apps are limited by memory bus speed. Some are limited by network speed. Some are limited by level 2 cache speed. Others are limited by CPU speed. It all depends on the app. Very few games are limited by disk I/O. Very few static page web servers are limited by CPU speed (or so I would hope...). And so on.
  • Still, I'd bet has 90% of the files they serve, in memory cache (switching seamlessly over to discussing FTP servers). And, doesn't that extra price tag justify getting another server entirely instead? As an added bonus, you'll have fallover support :-)

    /* Steinar */
  • by Anonymous Coward
    Isn't anybody worried that larger companies will try and use new standards to oust linux development? What did PCI do for us? winmodemes, no dsps and moving sound work to the cpu. All of which are not the most linux friendly things. Our fallback is the good old ISA. My old system used to have PCI everything but I upgraded to ISA everything (AGP video of course). What happens when even newer systems are out? Look at athlon boards - best processor available but the motherboards have 1 ISA ,just 1! If this is progress I will go back to my shanty.
  • Sure, dude. First, it isn't new. It's pretty new to the PC industry (although I heard of a guy who stuck his 286 laptop in the fridge, so it would run like a 386... :) but Cray did this a long time ago...

    No, it isn't for everyone, just for people who apparently need at least the fastest uni-processor speeds available. That's what this is good for. It isn't good for price/performance, and I haven't seen anything said about multiprocessing, (although you could cluster them, at least) but if you need a high MHz number in a box, this will give it to you.

    All CPU-bound apps should speed up. (I could probably encode MPEG audio in realtime, even with a less efficient algorithm... with the setup I have now, it takes 2-3 times the playing time of the CD...) But don't expect everyone to get this until it gets cheaper than a faster chip. Only get this when there are no faster chips! :)
    pb Reply or e-mail rather than vaguely moderate [].
  • It'll happen...never underestimate marketing!
  • If a computer were only as fast as its (sp) slowest component, then we wouldn't have any need for L1 or L2 or disk cache, would we? Engineers have been working for years to ensure that computers aren't brought to their digital knees by their slowest components.

    Sure, engineers have been trying to work around these problems for years, in order to minimize the effect of the slow components on overall performance. For instance, many of the design tricks for CPU design involve trying to mask memory latency (ie caches, load bypassing, load forwarding, etc). But the fact is that you can only do so much to minimize these effects. They still present a bottleneck, and still impact performance. That's why Intel is pushing Rambus, and Via et al. are pushing PC133 SDRAM - because memory is still a significant bottleneck.

    And caches as we know them today are only effective because much of the data used in present apps has a high degree of locality - ie the same few pieces of data get used repeatedly, so caches make sense. But the trend is heading towards more and more multimedia stuff. And multimedia (for example, video or audio) tends to have streaming data types - you process the data once, and don't use it again. Thus caches aren't all that effective for a lot of multimedia processing. In order to gain significant performance, memory bandwidth/latency problems need to be addressed.

    Of course, one problem is that I/O doesn't get the glory that, say, CPU design does, so you don't have as many people wanting to work on it. I/O has been the neglected child of computer engineering since the days of Seymour Cray (who had some great quote that I can't seem to remember :), and continues to be.

    Sure, it would be nice if everything was as fast as the CPU, but that ain't gonna happen

    Maybe not. But if you improved the memory speeds by 10%, I'd bet you get better performance than by increasing CPU clock by 10%.

    Increasing the processor speed can only gain you so much. Problem is, for a lot of apps, half of the work that needs to be done is I/O, the other half computation. So you get diminishing returns from increasing CPU speeds, because that 50% of the work from I/O isn't going any faster. Even if your CPU was 10^10000 GHz, if you can't get the data to the CPU fast enough, it doesn't matter.
  • video card cpu's dont run at 400MHz. they don't even break 200. The new 3dfx voodoos won't break 200. talk about getting your facts right. and yes under certain conditions your fast as processor wont keep up with your video card. I have 2 voodoo 2's and a 500mhz celeron. the processor still cant keep up with them. usually.
  • "Actually, an _equivalently clocked_ PPC G4 outperforms an Alpha."

    This seems like a good argument, but all processors cannot just be manufactured at the same clock speed. The design of the microprocessor not only affects how much work you can get done in a clock cycle, it also affects how high you can clock a processor. Alpha's are highly pipelined which if I understand correctly actually decreases (sometimes? always?) the work done per clock cycle, and the length of each clock cycle, resulting in higher Mhz.
  • True, but if anything those results will be slanted in Intel's favor. It's been widely speculated that Quake 3 - or any FP intensive app - won't be as "optimized" for the Athlon's parallel FPU architecture. So the fact that it still beats a Coppermine handily, to means, means it's either fast, or very fast. Either way, I want one :)

    And again, I say, so what? He's testing real world performance using real world setups. The bottom line is that Quake 3 runs faster on an Athlon with a GeForce as opposed to a Pentium. I don't really care why it does that, I only care that it does.
    "Some people say that I proved if you get a C average, you can end up being successful in life."
  • This coming from an avid overclocker:

    Kryotech may be pretty cool (no pun intended), but it kind of defeats the purpose of overclocking. Overclocking is to get more for your money, and if it has the side effect of giving you some bragging rights than all the better. But first and foremost anyone who overclocks is just trying to get more performance for less money. The performance:price ratio is what it's all about, and spending a shitload of money on Kryotechs cooling system is kind of counter-productive. Except for those aforementioned bragging rights of course.

    If you set up some insane cooling system like many people do, you still usualy at least break even with the cost of buying a processor of the speed you've overclocked yours too. For example, say you spend $40 for a Celeron 366, and overclock it to 550 mhz. You've saved ~$100 than if you bought a 500mhz Celeron (slower overall, and on a slower bus speed (66mhz)), and about $270 over buying a P3-550. So you've got $100-$270 to spend on cooling to either get it to reach 550 mhz, to make it stable, or just for a safety buffer. That should leave you plenty of money left over to buy some memory, a new hard drive, or to just put in your pocket.

    I've been running a Celeron 300a @ 450mhz for over half a year with an 'aftermarket' heatsink & fan and it's still perfectly stable at 2.0v. I've got a 366@550 at 2.1v and will be getting another 366 and a BP6 as soon as I get the $$. So what? Well my point is the dangers of overclocking are usualy worth it. I've got some nice stable systems that I spent a LOT less money on than if I bought the processor at the clock speed it's now running at, with pretty much the exact same performance. If it takes a year off my processors life, big deal. With the increasing demands of (mostly gaming) software these days, the hardware gets obsolite long before it's going to burn out, even if you overclock it.

    I'm not saying overclocking is for everyone, just for us smart ones :).

  • According to this article, the complete system runs about $2500

    Actually, $2500 buys a bare-bones system with only the special enclosure w/ cooling system, motherboard, and CPU. Add your own memory, controllers, drives, peripherals, monitor(s), etc.

  • Where;d you get your Celerons, did you test them yourself or get them pretested? I'm talk about the 366s.
  • You would think that the most respected source for this sort of tests on the Internet could get it right for once.

    I agree. I have read Tom's on and off for a while, and am continually annoyed at their seeming lack of concern for anything but games. It is a fact that people do things with computers besides play games.

    • I do not consider Quake, Descent, and the like a reasonable test of much of anything. Those benchmarks do not tell me how fast it can do number crunching, or how fast it can serve up html. I do not and will not use windows, so I don't care how many business WinStone(d)'s it does.
    • How about useful measurements of the system, like measured I/O throughput, Apache benchmarks, standard SPEC benchmarks, hell, even Seti@Home would be more useful.
    • Because of the benchmarks Tom's chooses, their evaluations are useless in comparing their hardware to non-Intel stuff. I'd like to stop supporting the Microtel duopoly as soon as possible, and I suspect other people would too. Doing so requires knowing just how good other hardware is.

    I have e-mailed Tom's several times asking them to use more standard benchmarks, each with no response or acknowledgement. Does anyone know of another, better benchmark site? One that compares heterogenous hardware, on unices, with more reasonable benchmarks? Anyone want to start a sitle like Tom's for Linux users? I mean, hardware vendors will often send you free "evaluation" hardware to play with. That could be a pretty good incentive. ;)


  • They are notoriously unable to cope with new processor architectures.

    Like the 1983-era benchmark I ran that said my K6-2 350 was 5600 times as fast as an 8mhz 8088. Well maybe, but not that fast.
  • by gashalot ( 5775 ) on Sunday November 21, 1999 @06:57AM (#1515239) Homepage
    We are always hearing about the new speeds of the processors and how much better/faster they are than the previous processor. However, with every new CPU and the accompanying motherboards the same system bus remains: PCI, ISA (being phazed out), and AGP. I understand that AGP is continually getting faster (1,2, and now 4x), but PCI and ISA aren't getting any quicker. Esp. with the new wave of gigabit ethernet cards hitting the market, even if you assumed that the CPU and disks could/would push a full gigabit, when does the PCI bus hit the wall? Are there new busses in the works for the system internals, or will everything eventually be powered by USB2 or IEEE1394?
  • I also read an article thurs. at Tom's about how you can rig you own over clocked amd processor. Obviously you won't hit the insane threshold of 1GHz, but hey, IMHO, every MHz. counts. It was pretty cool and quite indepth, but definitely not a project for anyone afraid of a fair amount of rewiring of their brand new processor.
  • by heroine ( 1220 ) on Sunday November 21, 1999 @07:06AM (#1515242) Homepage
    Well the benchmarks still give only a 33% improvement over the trusty 550Mhz Celeron. A A dual Celeron 550 would still give you more MIPS. For encoding massive amounts of mp3s and rendering video in parallel the dual Celeron 550 has held its mark longer than any configuration since its time.
  • No pun intended. :)

    Seriously, though, this is neat, but not for what it is. Rather, if a slow processor can be overclocked to 1 GHz, by supercooling, then upgrading the processor to something that naturally does 1 GHz, and THEN applying the same cooling, should give you an AWESOME speed!

    Besides, new processors aren't cheap. They can cost as much as much as a new computer, with the previous generation of processor on board. The cooling might work out cheaper, in the short term, for the same performance.

  • by Hobbex ( 41473 ) on Sunday November 21, 1999 @07:09AM (#1515245)

    In true Tom style, he goes ahead and sticks a geometry accelerating Nvidia GeForce card in the computer he uses to test the 3d performance of the processor (effectively testing the card rather than the processor). Tom has been nutorious for always choosing setups that create bottlenecks in the wrong places, in fact, his 3d-card tests are usually the best places to look for processor performance, and vice versa...

    You would think that the most respected source for this sort of tests on the Internet could get it right for once.

    Also, since this is Slashdot we are about to get a hundred posts saying something along the lines of: Processor speed doesn't matter, its X (replace x for "harddisk read/write","bus speed","cache memory" etc). Don't believe them. Yes, there are applications where these things matter more (specifically server activity in most cases), but for 90% of us the only applications where speed is an issue at all any more are the 3d apps where it is all about processor speed (except in the case of geometry acceleration as covered above).

    We cannot reason ourselves out of our basic irrationality. All we can do is learn the art of being irrational in a reasonable way.
  • With that sort of system, are they completely sure of correctness and stability?

    For instance, is it still inadvisable to run computational tasks where the arithmetic *must* be reproducible on other platforms, and where the system must be able to keep this up for, oh, a week while a job completes?
  • 3DNow is floating point. It is, however, hampered in part by the FPU architecture, because it uses the FPU stack for it's registers, IIRC. That means that as soon as you need to do something that 3DNow doesn't do, things get messy. Or so I believe, I haven't looked closely enough at it to be certain of this.
    3DNow also only supports single precision. I don't know what "guaranteed accuracy" means here. I seem to recall that the iterative square root and reciprocal algorithms can give you 23 bits of precision, though.
    This is all based on reading the 3DNow spec for the K6-2 a long time ago, so it might be wrong or it might not apply to the Athlon.
  • I would agree, actual MHZ is irrelevant. But game performance (to a point) is critical. When ya play competively, every little extra means something. I dont go out and buy the latest and greatest just to have it. i upgrade when i feel the ugrade will improve performance by a signifigant margin. I dont sell a p2 500 and buy a p3 550, but I will upgrade to a v5 6000 (from my tnt2 Ultra) if it pans out to be a good product. just like i will dump my dual 550's for a athlon 750 (when prices come down). and when adsl comes to town, i will consider dropping RoadRunner.

    If you play games online casually, this theory may seem rash. But, if ya take it seriously, (everyone has their hobbies, right?) then you have to keep your equipment up to date. to each his own.

  • Encoding mpeg (Sound and Video)? I would like a system that is able to encode mpeg video in real-time! But I think IGHz still isn't enought for that?
  • Except of course for those dual Celeron 600s. :)

  • I can't really understand why people want to go for 1 gigahertz instead of 900 MHz, when you have to pay that much more anyway :-) But I'm not sure if a dual Celeron would outperform this one; at least it would depend on what software you ran. Memory-intensive software, for instance, could easily `fill up' the bus, leaving you with only a _few percent_ left for improvement. That's a general problems with the Alphas too now -- the multipliers are so high, the CPU is spending more and more time just waiting for data... That's why Coppermine could be a relief -- going from 100 to 133 MHz helps a lot more that upping the multiplier by 33%. A bit funny that that is exactly the number you quoted, though ;-)

    /* Steinar */
  • Most normal webservers would probably don't care about access speed at all. Your entire company website is supposed to be small enough to fit in cache, unless you're running something with user input (like Slashdot) or are a fileserver (with lots of files). Generally, what is the most requested page on a website? The front page. So, naturally, it stays in cache all the time, and the hard disk will stay nice, cool and relaxed :-)

    As the person above you in the thread pointed out, RAID helps too, if you need it. Or just split the files among multiple disks (for a fileserver).

    /* Steinar */
  • OK, I'm kind of new to these super fast computer models (>800MHz) and I am completely amazed to see that they're using compressors to cool their chips. I remember about 7 or 8 years back hearing about the pentium design having the flaw of overheating. What a great fusion of technology (old and new) of having a refridgeration unit built right into your PC. Say goodbye to your heat sink, your "thin fin" cooling mechanisms. So a few years down the line when I finally buy my GHz home computer I will not only be buying myself a superfast computer, I'll be buying a refridgerator similar to the one that I keep in my dorm room. I didn't read the entire article in Tom's, so I'm wondering about two important things. A) How much more energy will my computer be using? Refridgeration systems cost money to run, will it be economical for me to keep it on 24/7? It will be like having two refridgerators in my home in use. B) How much will my new beer cooler/computer cost? I figure tack on an extra couple of hundred dollars for the refridgeration unit. Energy. We US'ers already comsume far more energy per capita than any other country in the world. Now with our new PC's that need refridgeration units, I'm sure our per capita energy consumption will skyrocket. Any thoughts?
  • If these comparisons largely depend on the graphics card (and possibly its drivers) instead of the processor, then you are not getting the whole picture.

    If these comparisons depended largely on the graphics card, then they would show the same results at all processor speeds. Which they didn't. Ergo, the comparisons don't depend largely on the graphics card. I'm not sure why that was so difficult.

    Now, it is true that in one specific case [], out of seven benchmarks posted (note that it was *not* Q3:A, which scaled almost linearly with the CPU), the video card was the limiting factor. Of course, that situation is (obviously) liable to happen, and if Tom didn't show that, then he wouldn't be giving you the whole picture.

    And, as noted before, the idea that anyone in their right minds would spend $2500 for an 80 pound supercooled computer and play games on it with anything less than the best video card is patently ludicrous.

    I guess that's why I'd rather get my benchmarks from Tom than from you.
  • is a site alot like you're talking about. It doesn't usually have server type statistics, but they do show how long a kernel takes to compile :)
    plus, the site admin reads slashdot.
  • I'd just like to thank Kryotech for naming it after me. I'll even let them buy my url for a cool million.

    This is a joke, albeit a poor one. Do NOT send any death threats, or intellectual property/copyright infringement lawyers.
  • AFAIK, the two reliability factors cancel out - the refridgeration helps the reliability, whilst the overclocking stresses it. Now if you were to overlock without refridgerating, as many people do, then I think you are playing with fire.

    Like another poster stated, I'm not very impressed with overlclocking in general - I mean if you were talking about 100%+ performance improvements then I think it would be worth the trouble, but if you're only getting 10-30% improvement over a normal system then I'd forget about it and go buy a multi-processor system or cluster.
  • Right now, it isn't the speed of the CPU that is slowing down computers... Nor is it the Bus (Most of the time). It's the HD access speed. HotlineHQ [], a compny I used to work for, is a great example of this. The slowest part of the whole chain on our webserver is the hard drive (Ultra-wide SCSI II). and the server is only a 266 (Granted, it runs normally at 75-90% capacity). Hard drive manufacturers keep insisting on Larger drives, not faster ones. So we have this enormous riverbed for information to flow through, but a dam up the stream is preventing the full usage.
  • I'd have to disagree on that.

    If you take an old PII board with SIMM slots, and you fill it with SIMM memory with a Pentium II 266.....You're going to get shitty performance. Given, PC100 RAM would perform better but for higher performance you would need to equalize the system out.

    If we had the processor and system bus running at the same speed, the computer's performance would be phenomenal. But since Intel is pushing processor clock speeds rather than improving the rest of the system, we have a Megahertz (gigahertz?) war between companies.

    A Computer is only as fast as it's slowest component.

    -- Give him Head? Be a Beacon?

  • Great, now I can take my motherboard out of the freezer [] and pound that vodka I was using get the cooling for those last few Mhz!
  • Simply adding fans to the case doesn't really do much for cooling the system, the idea is to get the air flowing through the case (usually with a fan in the front sucking in air and a fan in the back blowing out the now hot air). Running with the case off doesn't really help cooling much, the forced air from the fans dissipates heat alot faster than whatever breeze happens to be blowing through your computer room.

    The idea of cooling the whole case could be done quite easily by putting a coil on the intake fan, ofcourse this will only cool a few degrees and nowhere near the cooling power of the kyrotech unit. Once you start cooling the case more than than you need to worry about condensation building up -- not only could you short your system you could also rust it. (most people avoid this by having the cooling assembly well insulated from the rest of the machine and no air gaps for condensation to occur.) ofcourse you could always submerse your computer in non conductive mineral oil [].
    - MbM
  • Well, the PCI bus is divided to run at 33.33... MHz if you are running at one of the standard CPU/Memory bus speeds (defined by Intel to be 66 or 100, and now 133). Some boards have separate clock generators so that the PCI bus can always maintain the correct speed even if you are playing with your FSB at 75, 83, 112, 124, or anywhere in between. ISA is slower (8-12MHz) and nowhere near as fun, but PC PCI is still chugging along at a good old 33.3 MHz.

    Just my $(.004)^.5
  • I obviously expressed myself somewhat unclearly. Alphas are some of the slowest chips when you compare them to other CPUs running at the same clockspeed. What I meant to say was that Alphas have usually had the fastest clocks among the mainstream CPUs.

    But it looks like x86s are overtaking them. This is quite an accomplishment since an x86 can do quite a bit more work per cycle than an Alpha.

  • A Computer is only as fast as it's slowest component.

    <sarcasm>My god! You're right! I just plugged my old 1200 baud modem into my Athlon 550, and the whole thing slowed to a crawl.</sarcasm>

    If a computer were only as fast as its (sp) slowest component, then we wouldn't have any need for L1 or L2 or disk cache, would we? Engineers have been working for years to ensure that computers aren't brought to their digital knees by their slowest components.

    Sure, it would be nice if everything was as fast as the CPU, but that ain't gonna happen. That's why we have caches, pipelining, etc.


    Note 1: aargh! why does HTML preview change by &lt; to a <???

    Note 2: what does pipelining have to do with it? Pipelining lets the CPU (conceptually) run faster than its logic gates allow by doing more than one thing at a time.

  • Two things. First, Tom's site has always been heavily oriented towards the performance/gaming community. While he could litter his site with CPU and FPUmark bench tests until his heart is content, no one, myself included, would care. If I were to buy a G-Force, it would be for one thing - games. Thus, who cares about testing the CPU? I want to know how fast my GeForce/TNT2 will do in that machine, and that's exactly what Tom has told me.

    Second, the hardware bottleneck is in the CPU, not the graphics card, with the exception of the DMZG test. Performance scales somewhat linearly for every other 3D test.
    "Some people say that I proved if you get a C average, you can end up being successful in life."
  • by Anonymous Coward
    This is what RAID is for. Either split your web stuff into two disks, or put the whole thing on a striped/mirrored RAID array. Also, there is no such thing as UW SCSI-2. It's either UW SCSI-3 (3 implies wide) or Ultra2 SCSI (uses LVD and is 2x the speed).

"Regardless of the legal speed limit, your Buick must be operated at speeds faster than 85 MPH (140kph)." -- 1987 Buick Grand National owners manual.