Follow Slashdot stories on Twitter

 



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

Intel Issues Gag Order on Pentium III Speed 91

Matthew Tebbens wrote in to tell us that Intel has gagged vendors from saying how fast the new P3 processors really are and the Preview Day in San Jose on Wed. What are they hiding? We all know that its performance sucks on most stuff.
This discussion has been archived. No new comments can be posted.

Intel Issues Gag Order on Pentium III Speed

Comments Filter:
  • and the speed's nothing to get overly excited about. No matter what Intel says, the P3 is jsut a P2 with more instructions- same archtecture, same core, just more registers and more instructions. Which is all well and good if the extra instructions are really usefull.

    Is the P3's KNI useful? Well, kind of. There's no denying that using KNI you can speed up some kinds of floating point math (and only floating point- KNI doesn't do int). However, the speed increase is only really evident when doing a demo that heavily relies on that kind of math. When KNI is only a part of a much larger app (such as a game), which not only does 3D math but also lots of custom drawing, sound, AI, input processing, the speed increase is pretty much lost in the shuffle.

    So, in summary: My pre-release Katmai400 (or P3 as they're now called) is EXACTLY as fast as a P2-400. Don't anybody get too excited, it's just the P2 repackaged.
  • It's definitely not an inherent X thing that makes games on X slow. In fact, Quake is faster in X than on the console for me. And look at SGI. They do their graphics in X. Are you going to suggest that they are not fast?
    Now, what we need is a good OpenGL infrastructure (being worked on by Precision Insight) and good drivers. Unfortunately there doesn't seem to be much interest from the hardware companies in releasing either specs or drivers.
  • by drwiii ( 434 )
    The added speed and 70 new instructions make 3D games come alive and voice recognition sing.

    Um, ok.. That line convinced me to get one. .. .. NOT.
    Show me the numbers.

  • Posted by Lord Kano-The Gangster Of Love:

    You are gravely mistaken, the G4 also will sport a revamped FPU, and other internal architecture changes, like 128 bit data paths, and support for SMP.

    The G4 is to G3 what PPro was to P54c.

    LK
  • Posted by D. Black:

    I think it's more like Intel needs *something* to sell processors for. Let's face it folks: you don't need much more than a P-166 to use MSIE, Word, and Eudora. This is what is causing the rise of the sub-$1K computer (which is dominated by AMD, not Intel). Unless there's a new gadget or nifftiness, Intel's pretty much screwed. So they do nice things like develop "software" DVD decoders and tell people that they need new processors so they can talk to their computers, even though voice recognition isn't even close to being acceptable yet.

    On another note, has anyone noticed that Hotmail has slowed to a crawl? Microsoft must have moved it from Solaris to NT. I guess I'll just have to move my spambox to Yahoo mail or something.
  • Posted by D. Black:

    Realistically, how much more bloated can Win2000 (in terms of processor requirements) be than Win98? Remember that 2000 is now no longer based on the NT kernel, but on the 98/95/DOS kernel. Unless Microsoft decides to "integrate" MS Office and/or voice rec, I doubt it's going to be any more interesting than the 98 "upgrade" from 95. Besides, even if it is bloated, the price points for processors will be such by then that PII 450s will be in the sub$1K range because Intel is spending $300 Million to tell us that the PIII is oh-so-much better than the PII, which puts the PII into the Pentium category when the Pentium MMX came out. And I tend to doubt that even Microsoft can out-bloat a PII-450 without doing some really really funky stuff in the next two years. A PII 450, even when PIII 700s are out is still a very fast chip. Especially when all you want to run is Word, MSIE, and Eudora. And Intel can't make money on sub$1k computers. At least not enough to pay for a $300 million advertising budget.
  • dud you are making some pretty hardcore assumptions here.

    first off the reason thoes demos "scream" is that they take advantage of a new set of instructions. These instructions are designed for multimedia work and do infact accelerate certain tasks when software is written to take advantage of them.

    Similarly in the G4 there is a new peice of technology (altivec) which does very similar things, also requireing software to take advantage of it.

    The difference is that altivec is capable (theoreticaly anyway) of a FAR greater increase in processing power than KNI (the instructions in the PIII). I think you are being a little bit premature with your assumptions about the G4.

    Also, by this time next year, we should be seeing "multi-core" G4's with efectively 2 or 4 chips on a single processor, allowing for some serious speed. meanwhile you can expect the pentium arhitecture to begin topping out in its ability to increase in speed... youll notice that the frequency at which intell releases faster chips has decreased over the past few years... you wont see another substancial revoluton from intel untill IA-64 becomes available.

    in addition I would encourage you to look at these chips from a price performace perspective. despite a smaller overall market and therefore smaller volumes of chips the G3 is substancialy cheaper than the pentium II despite its exceptional performace.

  • Hey, lets load our CISC with more CRAP!
  • Graphics boards actually do do the geometry onboard, just not the lowend consumer grade cards.
  • Granted AltiVec is a juicy piece of silicon, and I personally cant wait for any package with such a specialized unit... however I am quite disappointed in the PPC triumvirates seeming inability to spell out a clear roadmap, or even come to a concensus on WHAT such fabled processors as the G4 and G5 will actually end up being... I can only think back to the major long-term hype of the PPC 620, and a few of its promised progenitors that never made the cut...

    The fact that the G4 is now a single-core, enhanced G3 (more or less a G3 with 604-style FPU from what I understand), plus *HOPEFULLY* the most-impressive AltiVec, all running on an interim bus that is supposedly the stepping-stone to the REAL G4 implementation leaves me feeling as unsure as ever...

    I also tend to think that we cannot defend the PPC platform based on any theoretical gains in efficiencies when the end-market prices end up being high due to the x86 platforms economies of scale... Intel will ALWAYS best PPC in market share as long as PPC is always within a questionable margin of performance... and with the collective engineering resources of the Intel flock you can bet your ass that they won't be caught too far offguard to lose their throne anytime soon... least not by anything less than a mindblowingly enticing incentive such as PPC doesnt seem ready to offer.

    I have heard the promise for YEARS that CISC and x86 are DEAD, but that fiction is based on the assumption that Intel doesnt adapt technologically... yeah their chips may be HUGE, power-sucking case-heaters, but they are a FAR cry from the x86 of yore, the line between CISC and RISC is nonexistant, more just an argument of semantics... SIMD is hardly CISC, and AltiVec is surely not RISC.

    Hmmmm... I love my PowerPCs, they kick some serious ass, i love the vindication of an alternative platform at least keeping pace with the mainstream offering, but dont fool yourself into seeing a great divide when there is none: my next box is x86, though Im sure I will never savor the dirty taste of Intel inside.

    Why are we so irrationally emotional over hardware (let ALONE operation systems or programming languages)... oh yeah... we're geeks. ;-)

    Binary Boy
  • Consider the following:
    • 21164-series Alphas have the highest cost:performance ratio(int and fp) of any non-x86 processor(that exists), far higher than Sparc and Mips.
    • 21264-series chips have thehighest performance of any cpu's, and still have a higher cost:perf than other commercial unix offerings. (dual cpu systems can be had for ~$15k)
    • Intel has "monopoly-like" influence, and has been seeking to extend that influence in ways that threaten our privacy.
    • (w)Intel's successes directly benefits Microsoft.

    Yes, I am a broken record.
  • Looking at the Chipmerchant price list [thechipmerchant.com] I see that the PII 450 is $547 and the PIII 450 is $550, a whopping $3 difference. In typical Intel fashion, if you go for the PIII 500 you pay $755
  • You see, MacOS cripples PPC performance. People have done off the cuff performance comparisons between PII's and G3's using Red Hat and Linux PPC. Integer performance is about on a par with the PII, floating point is faster on the G3.

    You really can't do side by sides using Windows and MacOS on these processors- it's comparing apples and oranges (no pun intended).
  • But FP performance is really only an issue if you're doing "Pentium" tricks- issuing an FP instruction and being able to out of order execute one or more integer ops while the FP unit was chugging on the request. More often than not, you only need these sorts of tricks on the rendering engine code- where 3DNow! has already been proven to shine over a stock PII at the same clock speed.
  • Read "P///" as "pee slash slash slash".

    ;^)

  • That article was some of the most disgusting, mindless uncritical hype I have ever read on the web. It makes all those "Is your computer ready for Windows 2000?" articles look positively objective by comparison. The "journalist" who wrote that drivel even tried to make Intel refusing to say what hardware their demos were running on into a plus-feature. This is a new low for computer journalism.
  • Comparing Photoshop in real-world tests on popular platforms is one of the only benchmarks that is truly meaningful.

    Jamie McCarthy

  • I do not want to have a conversation with a computer.
  • Dunno, I can't really see myself jealous of anyone talking to their computer or wanting the computer to understand them faster. I mean.. wouldn't it be ALOT easier to just go and get a girlfriend or boyfriend?

    *ahems* But then again, these days, exotic forms of intercourse are quite prevelant, aren't they?

    Yeah, but then again, considering the price of one of those suckers, one wonders if there isn't a "don't over use or you might go blind" warning label on the stick.

    ( a totally uncalled for post, but bored and well, if you really are waiting half an hour to an hour to have your computer learn your voice and commands, then well.. who knows. )

    ( yeah, that was uncalled for either. But some might even find it funny. )


    - Wing
    - Reap the fires of the soul.
    - Harvest the passion of life.

  • Flash! Flash!

    Shortly following the announcement and release of the AMY-3X, the current giant in the silicone idustry, Intehl ordered the gagging of all representatives and companies which sold or used their special P3 implants. The industry is in an uproar over this. Not just for free speach, but mainly because by god, that's was what they wanted to do at home.

    The AMY hasn't made any comments on their competitor's latest move, they are glad that they hadn't gone so far as to gag their employees. That is something that they thankfully do on their own.

    Compounded with their recent legal complications involving the monopolizing of the silicone hardware industry and the recently disovered peeping function of their P3, Intehl is definitely in hot water... no, not the gushing fun kind.. but more like the "it scorched my balls you @#$#%!" kind.

    In either case, the public, valued consumers of their silcone products and implants, are in a tizzy of the recent turn of events. In random interviews conducted just today, we can see that the opinions are somewhat inconclusive:

    ConsumerA: I think the new P3 rocks! My girlfriend is into this sort of thing and well.. I think it really helps us get work done, you know, especially how it helps to speed up the work involved with curves and modeling. I mean.. responsive hands on work of this kind was just not possible before a P3!

    ConsumerB: Well, all I can say is that if Intehl has to gag people, I want out. Maybe Amy is the way for me to go. I'm just not that kind of girl, you know? I mean.. sure, I want to be better endowed and get the faster and harder hitting Knky instruction sets.. but it's just all too expensive. I think Amy-3X with its lower price and good reviews will please me alot more in the long run. Besides, I don't like that gag reflex.

    ConsumerC: P3? Like.. what's that? *listens* Oh!! Yeah! I completely dig that! I mean.. I'm like using the P2 now and I get complete satisfaction every time, ya know? But I think the P3 will be the real pleaser for me.. I mean.. it takes alot of.. *listens again* Oh what? No way..! Damn.. where can I get my p3? Cuts down the oral dicktate time from 30min to 2-3mins? Hell, I'd pay $2k for that!

    ConsumerD: I don't know.. I mean.. where I come from, an 88 was good enough, you know what I mean? Didn't need all this Knky instruction sets and 3Xnow technology. I mean.. when our net-friends told us bout 369 and 369, we were shocked! I mean.. who needed all that extra stuff? It was just for getting work done, right? Well with all this P2 and P3 stuff going on, it's just too strange for me. I like to keep it simple.. one on the bottom and one on top. None of this dual and quad setups.. I mean.. that's just SICK!! I..

    Anyways, as we can see, the opinions are varied and well, to be quite honest, I don't know myself. I mean.. I don't mind dualing my systems, if you get my drift. But apparently the costs are somewhat high and alot of people are saying to just wait.

    But if you've got the money to go out and grab it, guess that's your kind of system.

    This roving reporter is signing off.


    - Wing
    - Reap the fires of the soul.
    - Harvest the passion of life.

  • Let me guess, you saw a video produced by MicroSoft that demonstrated this speedup?
  • It's a well known fact that Intel offers the services of their assembly wizards in order to create these outstanding claims of increased performance. Doesn't anyone remember when they were pushing MMX and quoting best case Photoshop filter speedups after their assembly team went to work on it?

    So, if this is an accurate statement, it probably means that the original code was not very efficient, not that the processor is somehow magically faster. I'm sure similar speed boosts
    could be generated if AMD went around with
    assembly guys optimizing stuff for 3D-Now!.

    I wonder how useful these new instructions really
    are. It seems like less and less people spend
    the effort to optimize code to this level. Unless
    the compiler does it for them the popular solution seems to be more memory/larger disk space/more expensive processor.
  • I don't know... everytime I see PIII I keep reading it as "Pill".

    I suppose this means that the K6-III can be read as K-Bill? Is that how much you have to pay for one? Or drop the "6" and it becomes "Kill".

    And all this before the first beer.
  • <lame_site_plug>
    Sorry, but I just had to mention this..

    Legions [legions.com] recently had a story [legions.com] about how Intel and their "advances" are not really pushing the envelope. It also goes into how AMD is a rising star in the processor wars and that Intel had better watch it's back.
    </lame_site_plug>

    Ok.. I'm done now. :-) Sorry.

  • > "There was no way to do this kind of operation reasonably before the PIII," Staiger said.

    So an extra 10-33% makes it "reasonable"? C'mon.

    > Graphics programs are clearly the biggest beneficiaries of the PIII's prowess, analysts said.

    Choice word "prowess". Which analysts said what?

    > The added speed and 70 new instructions make 3D games come alive and voice recognition sing.

    Now this hyperbole is getting ridiculous.

    Intel must put out pretty good lunches and drinks sessions to attract this sort of "objectivity" in journalism.

    -t. (should've become a tech journo)
  • Usually when that happens to me, I just drink some orange juice to wash it down.


    Oh, never mind.

  • I'm sick of seeing all these reviews and statements about the over-hype of the P3. The only hype I've seen is generated by the press saying how it's overhyped, or how it's not that great. Anyone who believed that the P3 would greatly enhance 3D didn' do their research on the instructions that have been added. The added instructions to help with 3D to some extent, but they are more geared toward speeding up specific algorithms, like the kind used in decoding and encoding video and voice recognition. In short, the P3 is offering the same type of idea as multiple processors: your computer won't necessarily be faster, but it will be able to handle more compicated calculations at one time. Thus, since the overall speed increase is minimal and mostly due to clock speed increase, and since NOBODY seems to develop the sorts of multimedia apps that would take advantage of these for linux, no linux user would benefit from a P3 over a P2. End of story.
  • We aren't really surprised by this are we?

    PeterT
  • First, let me say I'm clueless when it comes to 3d graphics.

    But I have to wonder:

    If the graphics calcs are parallelizable (and I thing that solving simultaneous equations is) then why not use a SMP farm as an "accelerator"? How about a large number of 486s or older 586 cores on the same chip or the same board? They should be able to pack these quite nicely these days.

    Since most of the hype is around graphics (ie. solving equations) why not just parallelize the damn thing? Any reason other than having to rewrite code so that it's parallelizable? For inherently parallelizable math code this shouldn't be too hard, is it?

    I know there has to be a good (technical) reason for the course vendors are taking, and someone here should be able to enlighten me :)
  • Mesa on 3DFX runs on top of Glide anyway, so if you can get Glide running with SLI... there ya go.
  • One positive I see coming out of this upcoming market blitz is totally unrelated to the actual clock or benchmark speeds of the Pentium 3 whatsoever...

    The price of top end Pentium II's will plummet, giving me a buying opportunity for "almost as much power as a P///"...

    Then I can wait it out for Merced.
  • This is a joke, right? Either that or a misprint. I mean, if they are comparing it to a "normal" pentium (MAX 166MHz I think), then they are obviously missing anything produced in the later half of the 1990s. This misses the MMX (MAX 233MHz), the Pentium IIs, and lets not forget the Celeron processors. Initial figures only have the P3 comming at ~7% higher on 3DMARKs than my Celeron (300A oc'ed to 450). Hardly anything to write home about. and certainly not worth the $1000+ markup in price for a P3. Of course I could talk for days about how MegaHertz (cycles per second) isn't worth the value that the industry tries to put on it. MIPS is the figure that counts.
  • My point was, that it's actually really dificult to compare any 2 CPUs. When you jump arcatectures i.e. P2 -> P3, you can't compare MHz. MIPS seems to be the best that I've seen, of course, what we really need is a standard Turing test to calculate minutes to complete specific tests.
  • I have a feeling that Intel is trying to get more hype going than anything else. From what I read about the PIII, it isn't going to perform anything greater, or anything worse than a PII. The only real performance gains you'll see starting out is the increased clock speed. Maybe SSE will become something, and maybe not, but starting out, I don't think it will really do anything.


  • by El ( 94934 )
    If your main focus is 3D gaming, then [the PIII] is going to be worth the investment...

    Uh, so I should really pay $700 extra for a chip that does what should have been done in the 3D accelerator board in the first place?
  • ...that the only ones being specifically gagged were the software makers. you see, their programs ran faster. much faster. and what does that best? prototype hardware! that is, i think intel took a bunch of the very best p3's they've made yet, ones that may clock in at >600, maybe put them on a 133mhz bus (if they have a vaguely functional camino yet) and cooled them way down.

    if the software makers can't comment, then they can't tell you that what you saw ain't gonna happen on anything you can buy any time soon.

    or i may be wrong, and they are just being perverse.

Sendmail may be safely run set-user-id to root. -- Eric Allman, "Sendmail Installation Guide"

Working...