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

 


Forgot your password?
Close
typodupeerror
×
AMD

AMD Athlon XP 2000+ Review 6 Weeks Before Release 175

Posted by timothy from the chips-to-beat-up-your-neighbor dept.
Mathew Solnik writes: "Tom's Hardware has a review of the AMD Athlon XP 2000+ 6 weeks prior to its official release. This review shows how to unlock the multiplier on the AthlonXP and how to reach AthlonXP 2000+ speeds easily." Note that by doing so, you are voiding any warranty you may have started with, risk blowing up your eyeballs, etc; do proceed with caution.
This discussion has been archived. No new comments can be posted.

AMD Athlon XP 2000+ Review 6 Weeks Before Release More

AMD Athlon XP 2000+ Review 6 Weeks Before Release

Comments Filter:

  • Please (Score:3, Funny)

    by cosyne ( 324176 ) on Monday November 12, 2001 @01:05PM (#2554236) Homepage
    someone explain to me why they had to choose that name.

    As far as I can tell, one product named XP is one too many.
    • The bandwagon is now boarding. All aboard.
    • It's called marketing. Remember the 386 ... 486 ... oh wait ... PENTIUM. It's all in a name.
      • Speaking of the name Pentium, isn't it about time for Intel to get off their ass and think up a new name? When they first decided to abandon their numbering system and not call their next chip the "586", the name "Pentium" was, admittedly, a clever one that implied the number 5.

        However, now that we've had the Pentium Pro, II, III and 4 I really think it's time for Intel to use a little creativity. Of course, part of the problem it that Intel was a little too clever for their own good and have now painted themselves into a corner. Sexium, Septium and Octium just don't sound as nice.

    • simple. they mean that you need at least this class of processor to make WinXP usable.

      at least that's my theory :)

  • I am just more and more impressed with the way AMD is going with there technology ... I just hope that keeping the same "basic" archtecture doesn't hurt them in the long run. I am looking forward to AMD chips in the future
  • ... considering that THG just had that infamous test like "what happens when the cooler dies?"

    And AMD's processors literally went up in smoke!

    No doubt you should have a fire extinguisher near!

    :-)

    • Re:Risky ... (Score:1, Interesting)

      by Anonymous Coward
      I make this simple for you. THG haked the motherboard that the AMD proccessor was sitting on. When a processor over heates the mother board shut off. By the way look at THG again and you see that the AMD proccessor fan was hooked stright in to the power supply, not in to the motherboard. So the motherboard could not recive information on the fan.
  • Note that by doing so, you are voiding any warranty you may have started with, risk blowing up your eyeballs, etc, do proceed with caution.


    We're all (most) adults here, there's no need to remind us of the consequences of our actions. If you put up warnings for this, you should put up warnings every time a kernel release story is out - "caution. This kernel may cause unstability, security holes, poor performance, oily discharge, etc".

    ~z
    • Software can be re-installed, but dead hardware is dead.

      If Slashdot doesn't print those warnings, could it be interpreted that they are encouraging behaviour that may physically destroy your system (and not "merely" cause downtime)?
    • Stat!

      I've seen those sort of tongue-in-cheek warnings on "new cool software" too.

    • OTOH, look at the number of liability lawsuits currently underway, particularly in the U.S.

      frankly, there are days when it seems to me that they should start teaching kids in elementary school how to write legally-binding disclaimer statements. not only for liability, but probably also for protection regarding ficticious works. recall the kids who have been punished for writing stories their teachers found offensive (i.e. relating to school shootings). was it appropriate? probably not. are there more appropriate ways to deal with this kind of thing? probably... but at this point, I'm waayyy off topic, so I'll shut up.

    • I would not have known that overclocking my processor would blow up my eyeballs, and I'm an adult. I'm glad that these warnings are in place. I'm also a humorless buffon that complains about sarcastic warning disclaimers.

      No wait, *I* knew that was supposed to be funny. I wonder who would be so dull as to think otherwise?
  • Another XP! It was bad enough when Windows went XP. Is AMD trying to associate itself with a software company out of Redmond?
  • So I don't see why this had to be posted at here. If you're interested about overclocking you computer you also prolly read those hw sites.

  • the good toms hardware (Score:5, Insightful)

    by NeoTomba ( 462540 ) on Monday November 12, 2001 @01:22PM (#2554341) Homepage
    Tom's Hardware has been getting a lot of bad press these days. Thier burning Athlon videos caused a lot of controversy, and regulars at some other tech sites (*cough* Anandtech *cough*) have a strong dislike for Tom's work.

    But this article just goes to show you why Tom's Hardware is the best hardware review site out there, just like it has always been. Over the years, Tom has written some of the best articles for hardware enthusiasts and has pushed "overclocking", a term which was once mysterious, into one of the big issues of modern computing. Motherboards are now being designed specifically for overclocking, and this has lead to huge increases in performance. Most people forget that Tom's Hardware has been one of the most influential sites as far as this is concerned.

    It's good to see what is undesputably the best current article on technology. I hope even those who hate Tom's Hardware will see the light. No other tech site has anywhere near the ability to do stuff like this.

    -NeoTomba

    • Re:the good toms hardware (Score:4, Insightful)

      by XBL ( 305578 ) on Monday November 12, 2001 @01:44PM (#2554449)
      Overclocking was cooler back in the olden days. Taking a 300 MHz Celeron to 550 MHz, or whatever was much easier and more productive.

      Looking at the benchmarks, doing this is NOT worth the time and effort.

      Tom has the time and money to dink around with these types of things. It might be worth it to him, but this article is likely worthless to 99.9% of his readers, simply because they aren't going to do it.

      Compare this with overclocking a Celeron, I bet over 50% of the readers gave it a try. Writing those articles were actually relevant. I'm sorry to say that this one isn't.
      • Overclocking was cooler back in the olden days. Taking a 300 MHz Celeron to 550 MHz, or whatever was much easier and more productive. Overclocking sure was much cooler back in the olden days. But a Celeron is "olden days"?! Now, if you had mentioned tweaking a 286, 386 or 680x0 (you get the idea) up 5 or so MHz, I would have agreed with you. Overclocking in those days wasn't "because it's easy", it was "because it's there"-- a much better reason to recklessly endanger your hardware in my opinion.
        • Yeah, but was overclocking even that well known back then? Was it well tested enough (like with the Celerons), that people would like to try it with some confidence... and if it screws up... no biggie buy a new Celeron for $40.

          Yep, that is the olden days, already. The BX platform is dead, and the old overclockable Celerons are basically long gone.

          I also think that overclocking will be less relevant in the future. Unless I can take my 5 GHz Athlon to 6+ GHz with relative ease, I'm not going to care much.
          • The one is a measure of quality, the other of time.

            Not that I ever saw a 286 boosted more than 4 MHz, but that was still a 33% increase. Testing? Who needs testing? :)

            Overclocking was never so fun after you started being able to do it all with jumpers (or, God forbid, in the BIOS). And once the Celeron 300A came along... well, it just lost all its mystique. Remember, geeks lose interest in popular things. I haven't overclocked since a brief stint with an old K5-100 at 150. (Very brief.) You mark my words, Linux will lose friends quickly if it ever sits on more than 10% of the world's PCs...
        • Back when I was a slip of a lad, apprenticed to a bunch of telephone engineers at IBM who thought PCs were beneath them, I "acquired" the clock chip from a 3750 PABX. Since that ran at 13.6MHz, and my PC-AT ran at 11MHz, I popped the case off the AT and switched the chips. Voila! 20+% better performance, but the 287 math coprocessor stopped working until I replaced the original clock chip.
      • It might be worth it to him, but this article is likely worthless to 99.9% of his readers, simply because they aren't going to do it.

        Mmm... I would call it valuable for a couple of reasons. One, it's geeky, so it has some interest to me whether I'm going to do it now, some time in the future, or never. Two, it's a good indication of just how stable a processor is, and how close the manufacturer is setting the "recommended" clock speed to its "maximum" speed.

        One other thing is, we're overclocking now by as many cycles as processors ran at full speed just a few years ago. Impressive or not, it's still interesting to watch people shrug at speed improvements they couldn't have imagined a decade ago.

      • Relevance of Publications (Score:4, Funny)

        by gmhowell ( 26755 ) <gmhowell@gmail.com> on Monday November 12, 2001 @03:53PM (#2555011) Homepage Journal
        A ferrari vs. lambo vs. porsche shootout is pointless, yet many magazines do them. How to build a $10,000 small block engine is pointless, yet there is at least one or two per year in hot rod.

        Many publications do pointless things. Not always is the point "this matters". Sometimes it is "isn't this some cool shit?"

        Like the Linux kernel source. I'm not going to monkey with it, but a handful of others will.
      • If you had read the article you would have understood that there are two points:
        - overclock the processor. OK that's geeky, not that much a performance gain compared to the old-days, but that's still funny. That's like this guy who put his PSX into a portable console. Not many people will do it, but that's fun
        - second point and that's why Tom's principaly did it, is to be able to test the XP 2000+ 6 weeks before it is out. Isn't that a nice thing to know?

        Even for AMD the news is good: it creates publicity 6 weeks before their processor is out, 'against' their will (didn't they protect the processor against overclocking?) and the risk of people overclocking the thing is small.
    • and has pushed "overclocking", a term which was once mysterious, into one of the big issues of modern computing

      No offense intended, but it's only a big issue among the script kiddies of the computing world. Everyone else just thinks "Hmmm...I could get a 10% higher clock speed, for a total system throughput increase of 2%, and there's the chance I'll either destroy and expensive processor or gain hard to track system stability problems. Or I could just let well alone. No contest."
    • No other tech site [besides Tom's Hardware] has anywhere near the ability to do stuff like this.

      You might want to look at icrontic.com [icrontic.com] or at this blurb about how to use trace tape to unlock the Athlon XP [plycon.com]. I know I saw the latter something like a week before I saw the Tom's Hardware article.

    • I remember running a 486SX25 at 40Mhz number of years back. So I'd say OC'ing was around, and popular, way before Tom or indeed the web.

      LEXX
  • I might just wait for a awhile. I am posting this on a new K7s5a with a 1GHZ AMD, MDK8.1 and this things is smoking already. I dono how much faster I could get!

  • Will people buy it? (Score:3, Insightful)

    by xdangavinx ( 534619 ) on Monday November 12, 2001 @01:22PM (#2554348) Homepage
    With the down turn in the economy, and there not being a great deal of software that non computer savy people use that require processors of such speed do people think that there will be a great demand for this chip in the consumer market?

  • Very nice ... BUT (Score:5, Interesting)

    by arminh1974 ( 530747 ) on Monday November 12, 2001 @01:24PM (#2554360) Homepage
    Now all that is quite an achievement, that they figured out all what's involved, but let's face it: Overclocking an Athlon XP 1900 to 2000 won't do anyone any good. That's like 2% more performance and at what kind of expense and risk? What would be informative and what wasn't provided would be if an Athlon XP 1500 (1.33GHz) can be rigged to reach 2000+ (1.66GHz) that way. It's all about how much headroom the CPU-core has and the price/performance overclocking provides.
    Tom's article shows that the Athlon XP clearly doesn't have a lot of it. We can expect the Palomino core to stick around the 2000-rating (1.66GHz) for a while ... at least until they go 0.13micron.
    • There were people posting about ocing their 1.33 & 1.4 Ghz Athlon XP's to 1.7-1.8 Ghz within days of the release...Aparently the smarter members of the ocing community have already proven the type of headrom a proper Athlon XP system has...

      AMD has been playing it very safe on clockspeed to make sure infrastructure is in place to get better performance (& not overwhelm or hurt next years profits due to performance beign to similiar between Athlon & Clawhammer). Tom is just an Id10t so he managed a whole speed grade increase with his oc... Wonder if he even bothered to use a non-standard heatsink... But I'm not about to read the crap he puts out to find out...

  • Connect 2 contacts. (Score:4, Informative)

    by laserjet ( 170008 ) on Monday November 12, 2001 @01:24PM (#2554361) Homepage
    I think it's cool that AMD has made it so easy to make this chip overlockable. According to the article, all you have to do is connect the L1 contacts, and that's it.

    although, keep in minde this is not for the faint hearted. you will also need a conductive lacquer to connect the contacts, tape, super glue, a scalpel, and multimeter. I don't know about you guys, but I think when all this is required to overclock your CPU, it's a bit much for the rewards you get.

    Also, for most users, the faster processors like this new AMD are so fast anyways, that overclocking them will probably give a faster experience only in benchmarks, and not in real user experience. It's a cool article, none the less - but if you just bought one of these babies, would You want to pull out all these tools to overclock it?

    I would play it safe and be happy with my already fast computer.

    • Re:Connect 2 contacts. (Score:1, Interesting)

      by Anonymous Coward
      The issue here is (for me), what is your time worth. It probably takes an hour to make this modification and you risk trashing the processor. With all that fancy modifying, you wind up with a few percentage points increase in benchmarks.

      Now, if you could take the very cheapest XP and achieve the same results, I might be interested....like the old celeron 300a that would clock right up to 450mhz by simply changing the bus speed from 66mhz to 100mhz. Those were the days. 8-)

      It is pretty interesting that the high end athlons spank the 2ghz P4 in most benchmarks and cost about 50% as much (and you get to use geek friendly DDR memory instead of paying through the nose for evil RDRAM).
      • (and you get to use geek friendly DDR memory instead of paying through the nose for evil RDRAM).

        Well, I don't know about the RDRAM, but my DDR [ddrfreak.com] memory is terrable. Maybe it's that I don't practice enough, but I can only keep one or two arrows in my memory at once. I suppose that it is geek friendly though, most geeks that I know are too embarrased to ask a girl to dance...
    • Easy? The amd engineers made it much harder this time. This is 100x harder to do than just connect them with a standard graphite pencil, like the T-Birds. AMD does not want you overclocking their chips, why do you think they cut the L1 bridges in the first place
      • It is not that hard to do. I think AMD doesn't appreciate OC'ing, and is making it EASY to recognize OC'ed procs (since you can no longer use a clean pencil line). They're making it hard enough so it really requires effort. And I bet they can tell when you superglue the holes and try to RMA it ;-)
      • The amd engineers made it much harder this time. This is 100x harder to do than just connect them with a standard graphite pencil, like the T-Birds.
        Actually, AMD's engineers didn't do anything specifically to make overclocking Athlons harder. Converting from the hard ceramic processors to newer organic grid array, for economic reasons, causes the laser cutting of the bridges to make deeper pits. It looks like they probably used the original laser power initially, since the first batch of Athlon XPs had large, charred pits, while newer batches have smaller, cleaner pits.

        <humor>I just hope they use soap and deoderant to clean their pits.</humor>

  • Nice!! (Score:2)

    by tcc ( 140386 )
    that means even more room to overclock when it'll shrink to .13 microns.

    I didn't think the current process would go above 1.5 with standard cooling, this is good news :)

    Now where are those Nvidia boards...
    • *cant find the article*
      But Intel engineers said that they have inspected the new Athlon XP and its made with .13 allready.
      (Even if AMD says they didnt.. )

  • hmm (Score:3, Funny)

    by sinnyin ( 530106 ) on Monday November 12, 2001 @01:44PM (#2554447)
    shouldn't that read "from the chips-to-HEAT-up-your-neighbor dept."?

  • How much difference will this make? (Score:3, Interesting)

    by Henry V .009 ( 518000 ) on Monday November 12, 2001 @01:51PM (#2554457) Journal
    I took a look at the benchmarks that Tom provided. Is anyone really going to notice the performance difference of overclocking their 1900+ to 2000+?

    It's a few hours of work besides, and they run the risk of destroying a piece of expensive hardware to do it.

    This space for rent.

  • Maybe I can buy an AMD chip (Score:4, Offtopic)

    by Anonymous Coward on Monday November 12, 2001 @01:58PM (#2554483)
    with my homemade "equivalent" dollars. Really, Mr. Dealer, one MyDollar® is the equivalent of $3USD.

  • Illegal (Score:5, Funny)

    by Anonymous Coward on Monday November 12, 2001 @01:59PM (#2554490)
    This sort of circumvention is illegal under the DMCA.

    In addition, since it makes Ahtlons explode it can be construed as a terrorist act under the recent counter-terrorism act.

    In short, Tom will be shot at dawn.

  • Be very careful if you do decide to overclock. (Score:5, Interesting)

    by Anton Anatopopov ( 529711 ) on Monday November 12, 2001 @02:00PM (#2554492)
    If I could give one piece of advice from personal experience, the pink thermal pad supplied with certain heatsinks is not adequate for the job. If you intend to overclock your cpu, scrape off all the pink crap, and use some proper thermal compound like arctic silver. Spread it very thinly, too much and it will act as an insulator.

    Also, bear in mind that not only will your warrenty be void, but some people have said you may be liable to penalties under the DMCA, since the clock multiplier lock is considered a form of 'encryption' and the increased processor speed gained by unlocking it can be seen as 'copyrighted software'.

    Quite how this can be the case is beyond me, surely the speed at which I run my software is down to me, but you never know with these DMCA issues. It can all get a bit surreal at times.

  • superglue (Score:4, Informative)

    by mc2Kleen ( 190152 ) on Monday November 12, 2001 @02:00PM (#2554494)
    If you read through the article, it points to several pictures where things went awry. One such example is where the superglue ran under the scotch tape onto the contacts and couldn't be scraped off.

    Some superglue manufacturers offer a thicker type that doesn't run quite as eagerly as the liquid type. It is more the consistency of model airplane glue so you have more control as to where the glue actually goes. It can be found at hardware stores and any place that offers a decent selection of adhesives. This may be a better solution than hoping and preying that your masking job is adequate and liquid superglue does seem to have a mind of it own sometimes as my fingers will testify.

    • How about skipping the superglue altogether? (Score:4, Informative)

      by Svartalf ( 2997 ) on Monday November 12, 2001 @02:18PM (#2554588) Homepage
      Since you're going to the trouble to buy silver laquer (in either the bottle like Tom's used, or in an applicator pen) you might as well go to the trouble of buying conformal coating material in a bottle or applicator pen- it's not THAT much more expensive. While it's cure time is a lot longer than superglue's, it's designed for this sort of thing and it's at least a little easier if you fsck things up with your end run around AMD's overclock blocks (because it's laquer and will give you some options to carefully scrape any overflow off of the lands.).

      By the by, this all seems like a lot of effort for little payback. Some of you may not want to do this.
    • Some superglue manufacturers offer a thicker type that doesn't run quite as eagerly as the liquid type. It is more the consistency of model airplane glue so you have more control as to where the glue actually goes.
      When I read the article, I wondered why they didn't just cut out a smallish piece of tape and use that to cover over the holes. A sliver of electrician's tape ought to stick to the processor package fairly well and would keep the conductive ink out of the holes. You could then use Scotch tape over it to mask off the pads for the conductive ink; I would think that the edge of the electrician's tape would be no problem for the ink to cover. It would seem to be much more foolproof than trying to plug the holes with glue.

      Another possibility might be to just route the ink around the holes...it's a bit tricker than making a straight line, but steady hands, a fine-tipped paintbrush, and a magnifier ought to do the trick

  • Well, duh!

    That was my first response.

    Now consider that Tbirds and P4's produce the same "horsepower" or have the same "oomph" despite one being 2Ghz and the other 1.4+.

    Ok, with that in mind, don't forget that the acutal size of the processor. The actual chip part of an athlon is, what?, 1/4 the size of a p4?
    Not only that but the p4 has a heat spreader (or first stage heatsync?) where the Athlon does not (almost typed in doe snot..heh, I love typoes).

    I'm sure thermodynamics ~= a simple physics question:
    Which exerts more force an elephant with a foot that has a 6" radius or a 100lbs female in high heel shoes (down, boy) with a 1/4" wide heel?

    The hinting was at which would hurt more, in essence. It went against most ppl's intuition.
    More force (or pain.) would be delivered by the 100lbs female on that 1/4" surface area.

    Similar reasoning applies to the Athlon. All that heat, on that small area. Did not help that the thermal shutdown sensor on the MB's did not poll quick/good enough (maybe the MB's were made in Florida, dunno. Cheap shot, sorry).

    Funny thing is this: if you did overclock, most likely you would leave the case's side off and would notice. And it is getting to the point with heat syncs that either liquid cooling systems are going to be needed soon, or anchoring it to the MB, a la p4's, is the next step.

    Either way I am going to build an AMD system soon before prices go up...only drawback is how to muffle the sound of dual 7k rpm fans w/o putting the thing outside.

    Cheers ppl.
    • I think you are pointing out that CPUs are producing more and more heat per square inch these days. However I have also noticed that CPU coolers are becoming more and more sophisticated as well.

      I bought a HSF (heatsink/fan, a GlobalWin FOP32-1) for my 1Ghz Tbird a year ago that worked well but was very noisy. I recently replaced this with a new cooler (an Antec JetCool) which is equally efficient and much quieter. I have seen a new emphasis on quieter cooling with a number of other products too.
    • "Which exerts more force an elephant with a foot that has a 6" radius or a 100lbs female in high heel shoes (down, boy) with a 1/4" wide heel?"

      The elephant foot exerts a lot more force. The heel exerts more pressure (even when you account for the fraction of the weight resting on the toe).

      Interestingly, if you work it out, the elephant's foot places about the same pressure on the ground as an automobile tire.
  • I can't wait to see someone show up for work with processor stuck to their fingers.

    I'm sure the song by Huey Lewis "Happy to be stuck with you" will surge in popularity for a brief moment.
  • Connect the pins with a pencil mark? Use superglue as an insulator? Normally I would shy away from attempting such a hack, but those mangled processor shots make me think - YOU CAN DO IT!
  • Anyone ever see any parallels between rabid overclockers that pay more time and money than it's worth to increase speed, and rabid automobile enthusiasts that pour money into building customized engines that they'll never be able to really utilize?

    To be blunt, there's some dickwaving going on here, as always. 1900 to 2000? Big whippity whoop, especially considering the effort and cost dangers.

  • Inadequate testing (Score:4, Informative)

    by Animats ( 122034 ) on Monday November 12, 2001 @02:31PM (#2554659) Homepage
    I'd be more impressed if, after their rather tacky jumper insertion, they ran a good hardware diagnostic program [ami.com] for 24 hours straight and didn't detect any problems. The overclocking crowd tends to think that if the machine will boot up, it's working. They're wrong.

    Tom's Hardware is also using the wrong tools for the job. Even other overclockers [mac.com] know better. There are pens for writing PC board traces with conductive ink. Using conductive paint and tape is doing it the hard way.

    • There are pens for writing PC board
      My limited experiece with conductive ink makes me think the stuff woulden't work too well on a searing hot processor. It doesn't adhear to smooth surfaces too well that are under stress - a circut board would be fine, but a processor that goes though heat cycles, would make me think that it would flake off over time. The Mac people probably get away with it because the PowerPC processors run cooler than the electricity hungry Athlon. Just my $.02 ($.01 after taxes.)

    • Re:Inadequate testing (Score:1, Interesting)

      by linzeal ( 197905 )
      Well I for one and most people I know who frequent places like HardForum [hardforum.com] the forum sister site to HardOCP [hardocp.com] use burn in programs like SiSoft Sandra's burn in program [demon.co.uk] for at least 12 or so hours before we leave the overclock in place.
    • This is a VERY small overclock. It is rare that a processor can't handle that. The only time I've seen a problem with relatively tiny OCs is in situations like the pre-tualatin P3's as you pass 1GHz (you might remember that they weren't even stable at their rated speed).

      But, I do agree a few days of non-stop testing is required before you can make an accurate judgement.
  • I read the article and found it interesting, but I fail to see why most people would want to spend the time and effort to do this. Why not just wait a few months for the next generation of CPU's to come out?

    Also, how reliable would this overclocked CPU be after a few months? What if the superglue or L1 contacts overheat and breaks a connection?
    • An hour's work against months of waiting. Maybe you're just being lazy.
      • Actually, it would take much longer than an hour to do that. You would have to drive to one or more stores to find the necessary items (super glue, silver stuff) and spend your hard earned cash on those items. After following all of those steps your CPU might just be destroyed if you make a mistake. If it does work, you are only looking at a 2%-5% increase in performance. This isn't a very realistic modification.
  • From the article:
    However, the maximum setting is currently limited to 12.5X, which allows you to reach a clock speed of 1666 MHz (12.5 X 133 MHz = 1666 MHz) without having to increase the front side-bus clock speed

    Or maybe what is being said is that the Athlon XP's are wickedly fast

    You be the judge

  • "Crack the processor?" (Score:1, Funny)

    by Anonymous Coward
    "we managed to prepare a clear, step-by-step guide so that all you ambitious PC users can crack the Athlon XP processor"

    I didn't need any help doing that. I've managed that my self before!

    "we also show benchmark results that demonstrate the jump in the performance of the Athlon XP/MP"

    Mine performance jumped right into the trash can after I "cracked" it.

  • What about the FBS? (Score:2, Insightful)

    by 10Ghz ( 453478 )
    I would have liked more results with overclocked FBS. If I got XP, I propably wouldn't bother to connect the L1-bridges just to gain few % of additional performance. I would increase the FSB. Easier, and it gives you more performance.

    I think XP is just begging for more FSB. Cranking it up to say 300Mhz (2x150MHz), would increase the actual MHz of the CPU, and it would nicely increase you bandwidth (both memory and FSB).

    Of course, I would much rather have the Clawhammer with it's 800MHz FSB...

  • Very, Very, VERY old news... (Score:4, Insightful)

    by frank_adrian314159 ( 469671 ) on Monday November 12, 2001 @04:24PM (#2555158) Homepage
    Overclockers.com [overclockers.com] had the information out a couple of days after the XP came out. This article [overclockers.com] came out on 19 October with essentially the same information as Tom's. Since then, at Overclockers, there have been other articles dealing with other approaches to reconnecting the bridges and how the laser cuts have changed.

    I don't know why people think that Tom's is a particularly good source anymore. These days they really seem to be slow off the mark...

  • I find it amusing that Tom's Hardware went through "
    several dead processors" before they figured this out. Especially since other hardware sites had posted this a week or two ago. Guess they had to read the articles and find out how to do it.

    Check out VR Zone's method. Much better and reversible.
    http://www.vr-zone.com/guides/AMD/AthlonXPUnlock/ [vr-zone.com]

    Tech Stats
    http://www.oc-athlonxp.com/bridges/ [oc-athlonxp.com]
  • So - can the XP be turned into an MP? Is this also just bridge work, or does it just drop right in?

    I'd kill to have a Dual 2000+ MP system...

Slashdot Top Deals

The key elements in human thinking are not numbers but labels of fuzzy sets. -- L. Zadeh

Close