Apple G4 90% faster than P4 2.53Ghz?

New benchmarks on the Apple website have some very ballzy performance benchmarks under PhotoShop 7. The benchmark shows a dual 1.25Ghz G4 system running 90% faster than a Dell with a 2.53GHz P4, and not only that the dual 867MHz system is shown as 45% faster. Clearly once again megahertz just doesn't matter, but 90%? LOL. The Dell 8200 w/ the 2.53GHz processor is Dell's current fastest system. The CPU had 512K L2 cache and ran on the newer 533FSB, as well used RDRAM PC800.

Check it out for yourself <A HREF="http://www.apple.com/powermac/specs.html" target="_new">here.</A>

"Opinions are like assholes, everyone has one"
32 answers Last reply
More about apple faster 53ghz
  1. An important question: How much does the Mac cost vs. the Dell?

    EDIT: I went and found Apple's price for the dual 1.25GHz. It is five grand! I also notice that the Dell's specs besides processor are not given. I went on their site and a 2.53Ghz with 512MB of RAM is around $1,800-$2,000. If you want to spend an extra three grand to make Photoshop run almost twice as fast, then be my guest.

    --S.<P ID="edit"><FONT SIZE=-1><EM>Edited by sdausmus on 08/14/02 02:08 PM.</EM></FONT></P>
  2. For the same price you can have a 4 way itanium 2 with terabytes of memory.

    The day i meet a goth queen that tell me Intel suck.I turn in a lemming to fill is need in hardware.
  3. apple computers have always been more expencive, they've also shown to have better hardware too. Remember, mostly people who edit photo's and video use mac's. Otherwise their just cute and small paperweights that check your e-mail =)

    "What kind of idiot are you?"
    "I don't know, what kinds are there?"
  4. LOL very true! I noticed the lack of specs also, but went to the Dell site and noted that's the best offering and featured PC800. Other than that the Dells specs are a mystery which doesn't say much for apple. And my god yes, apple is a very experience alternative. Talk about price fixing! Can anyone explain how they're so much more than a PC when most parts are essentially the same except for the processor and OS?

    "Opinions are like assholes, everyone has one"
  5. Please keep in mind that all Processor Manufacturers are biased on there benchmarks......they all "cheat" so to speak...therefore the only true place to read and BELIEVE benchmarks of the osrt r from 3rd part sorcess...like Toms, Anandtech, Ace's, and even some i dont know about...

    i did see a review on a Dual G4 1GHz system, a P4 2.53GHz system, Dual Xeon 2.2Ghz System and XP 2200+ system and Dual MP 2100+ system........BTW the Dual MP's took the cake overall.....i'll try and find the link...

    <A HREF="http://www.anandtech.com/mysystemrig.html?id=13597" target="_new">-MeTaL RoCkEr</A>
  6. I'd beleve that a dual 1.25GHz G4 Mac could beat a single 2.53GHz P4 in <b>Photoshop</b>. Beating it by 90% sounds awfully suspicious, though not entirely impossible.

    Photoshop (the one and only program that I know of that actually runs better on a Mac than on a PC) is <b>extremely</b> optimized for a Mac. (And still debated if it is an intentionally bad port to the PC, or just done by unskilled monkeys.)

    In any event, the reason for the severe performance difference is simple. G4s have a Level 3 cache (I believe something like 1MB). Photoshop is written to make <b>extreme</b> use of this. As such, the PC port of Photoshop is very sloppily done. To my knowledge (from what I've read) instead of re-writing the portions of the code for the PC port to do away with the excessive use of the L3 cahce that a Mac has, the code is more or less left as is, so that the PC port keeps trying to utilize a 0 byte L3 cache. As you could imagine, that results in a lot of un-necessary attempts to check, load, unload, etc. the completely non-existent L3 cache.

    Any <i>normal</i> programmer would have just removed all of that wasted code for the PC port and saved countless cycles of CPU time executing completely pointless code. Any <i>good</i> programmer would even go one step further and utilize methods to optimize the performance for an x86 architecture, possibly even optimizing for a P4. (For which there are several easy optimizations that can be made.)

    Since nearly every other piece of software that has been run on both a PC and a Mac has been able to be optimized for each seperate platform and run at reasonably similar performance (that is, performance in relation to the performance of each platform's CPU, not equal performance to both platforms regardless of CPU) it remains very suspicious that Photoshop is the one and only testement to the Macintosh's "superior" (not in my book) performance.

    I also suspect (though this is only a guess and rumor) that if one looked at the code, they might even find that the PC port of Photoshop also includes numerous (if not intentionally-excessive) bitshifts and other operations which the current P4s handle rather poorly. I would have to see a benchmark of Photoshop run on an Athlon to judge that better though.

    If one wants a decent evaluation of a Mac's actual performance compared to a PC's, then one should use software other than Photoshop, as that piece of software is, as stated, highly suspect.

    If I were using Photoshop exclusively, I would most certainly want a Mac. If I were using, well, <b>any</b> other software, I'd use a PC.

    P.S. And, of course, once again, one is left to only guess and wonder how the PC for Apple's comparisons was actually configured. The simple fact that they never say outright what the complete system specs are leaves much to be guessed about the actual validity of their benchmarking.

    <A HREF="http://www.nuklearpower.com/comic/171.htm" target="_new">The corpse you find may be your own.</A> - Black Mage
  7. Um, you can get a dual Xeon or dual Athlon MP system, and 2GB or more of RAM for the same price as that MAC system! For sure, that'll be faster than the dual G4. Who cares anyway, in this day and age, with enough RAM and a fast hard drive perhaps in a RAID, Photoshop will be fast enough with a top-of-the-line processor. However, video editing is still a demanding task on any system.

    :wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
  8. Quote:
    In any event, the reason for the severe performance difference is simple. G4s have a Level 3 cache (I believe something like 1MB). Photoshop is written to make extreme use of this. As such, the PC port of Photoshop is very sloppily done. To my knowledge (from what I've read) instead of re-writing the portions of the code for the PC port to do away with the excessive use of the L3 cahce that a Mac has, the code is more or less left as is, so that the PC port keeps trying to utilize a 0 byte L3 cache. As you could imagine, that results in a lot of un-necessary attempts to check, load, unload, etc. the completely non-existent L3 cache.

    Any normal programmer would have just removed all of that wasted code for the PC port and saved countless cycles of CPU time executing completely pointless code. Any good programmer would even go one step further and utilize methods to optimize the performance for an x86 architecture, possibly even optimizing for a P4. (For which there are several easy optimizations that can be made.)

    Long quote, short reply: Isn't cache-handling done by the hardware? No software ever needed to be recompiled or recoded in order to be ported from a 256K L2 cache-machine to a 512K cache machine (same architecture, of course, i.e. not from Mac to PC), I think. But I do agree that optima code will differ widely if you run on a system with or without L3 cache.

    Greetings to you all,
    Bikeman

    <i>Then again, that's just my opinion</i>
  9. I agree with you and would like to expand a little...

    Many many "professionals" are now actually using p4 (or even athlon) systems in order to do video editting (not photo, but video). Why? Because the video compression on the p4 (and to a lesser extent athlon) is dynamite and more than competes with Macintosh. In addition, there are quite a few more options for great video cards for PCs as opposed to Macs (my friend who makes wedding videos - using a duel althon 1.4 - just bought a 3 thou $$ video card...has about every type of compression built into the hardware itself). And... PCs are definitely cheaper machines (not to mention better for upgrading).

    In regards to performance, Macs and PCs are right together. In fact, more often than not PCs will dominate in many software applications (even though Macs may do better with Adobe Photoshop - photo editting, the P4 is optimized to perform better with Adobe Premiere - movie editting -- isn't that just funny), so I still have no idea why people are buying Macs. *shrug*. The world may never know...

    Athlons and Pentiums are just melted rock. Who’s rock is better? Who cares, let’s play some games
  10. A musician friend of mine says that the PCs crash all to often and that makes you lose track of what you were thinking, not to mention losing your scores in the software.

    Macs don't crash. Or so she says. I have limited experience of macs, but not unpleasant. So, I really can't comment except for they don't accept a lot of the latest hardware, don't have as many games and are unnecesarily expensive.

    EDIT: this was all a couple of years ago so...

    <b><font color=red>I'm a bomb technician. If you see me running, try to keep up.</font color=red></b><P ID="edit"><FONT SIZE=-1><EM>Edited by HolyGrenade on 08/15/02 00:49 AM.</EM></FONT></P>
  11. OS X crashed quite a lot, actually. I don't know about Jaguar, mostly because the people I know who have Macs aren't really keen on spending $129 for the upgrade, even if it is "about $1 per new feature."

    --S.
  12. I guess she has never used Win2k or WinXP.

    To start press any key. Where's the "any" key? --Homer Simpson.
  13. Personally, I am highly skeptical of anything Apple has to say about the performance of their computers. I have seen performance tests from PC Magazine comparing PCs and Macs. What tests the Macs could actually complete, they performed far worse than their PC competitors in everything except Photoshop (even then it still didn't perform as well as Apple said it would).

    Knowledge is the key to understanding
  14. My thoughts exactly....which i posted on already =)

    <A HREF="http://www.anandtech.com/mysystemrig.html?id=13597" target="_new">-MeTaL RoCkEr</A>
  15. Quote:
    Long quote, short reply: Isn't cache-handling done by the hardware? No software ever needed to be recompiled or recoded in order to be ported from a 256K L2 cache-machine to a 512K cache machine (same architecture, of course, i.e. not from Mac to PC), I think. But I do agree that optima code will differ widely if you run on a system with or without L3 cache.

    My understanding is that Photoshop includes a vast amount of low-level code to utilize the L3 cache directly. It kind of makes sense since the L3 would be used in a more 'generic' sense otherwise. In image editing software you could load large amounts of image data directly into the L3 cache for extremely fast processing.

    I'm sure that this code is engineered to not rely on any specific cache size, but simply to reserve as much of the L3 cache as it possibly can. However, I'm not so sure that the code is engineered to run intelligently were there no L3 cache at all.

    In fact one of the more evil rumors that I've heard (and again, I stress that this is a rumor, though the performance benchmarks do support it) is that the PC port of Photoshop uses so much of this low-level control that it was actually coded to use a 'virtual' L3 cache if there is no actual L3 cache. And since the L3 cache is accessed directly instead of allowing the CPU to handle that part, that means that the 'virtual' L3, being system memory, is being loaded and unloaded from other system memory. Could you imagine how many CPU cycles would be wasted waiting for memory transfers were the 'virtual' L3 cache concept actually being used? You'd basically just be doing a completely useless copy of data from one spot in system memory to another spot in system memory just so that you can perform your calculations, and then afterwords restoring that other spot's memory back into the original section of system memory. Sure, the software would work that way, but darn it'd be slow.

    So anywho, that's the speculation on just why Photoshop runs so piss-poor on a PC compared to a Mac. Not having seen the source code myself, I can't verify that any of it is really true. However if you look at the performance of other software on both platforms and then look at the performance of Photoshop on both platforms, it makes the rumors very believable. There is just no <i>good</i> explanation for the extreme performance differentiation.

    <A HREF="http://www.nuklearpower.com/comic/171.htm" target="_new">The corpse you find may be your own.</A> - Black Mage
  16. Against a DELL, I think my grandmother's hearingaid would outrun a DELL.

    I mean how sad is that they're comparing a cheap parts, mass production setup to their top of the line one?

    Also did they compared it with one software only? That would be pretty weak.

    <font color=red>Got a silent setup, now I can hear myself thinking.... great silence</font color=red>
  17. Yahoo! I got the funky language again. (I'm assuming it's either German or Dutch by the looks of it.)

    Quote:
    Udfyld felterne herunder for at skrive et indlæg til forumet. Det er ikke muligt at bruge HTML i dette forum. Tekstformatering er muligt, så det er tilladt at bruge tekstformatering i dine indlæg.


    He he he he.

    Oh, sorry, back to the reason I posted in the first place...

    Quote:
    <font color=red>Got a silent setup, now I can hear myself thinking.... great silence</font color=red>

    Heh heh. Who needs a silent setup when you can just get a monster 5.1 surround-sound speaker system and make ten to a hundred times more noise than your fans? ;) You'll never worry about your PC's noise level again. Ha ha ha ha ha ha ha ha ha!

    Yes, shoot me. I'm going sane. The medication that the government has placed into the water supply must finally be taking hold and driving me to sanity! Before you know it, Bush Jr's speaches might finally make sense to me! <b>Nooooooooooooooooooooooooooooooo!</b>

    (Hmm ... I wonder if one should pronounce that as an extended no, or as an extended nu on the basis of a double-o, as in moo. I originally intended the former, though the latter has a certain appeal...)

    <A HREF="http://www.nuklearpower.com/comic/171.htm" target="_new">The corpse you find may be your own.</A> - Black Mage
  18. Quote:
    Heh heh. Who needs a silent setup when you can just get a monster 5.1 surround-sound speaker system and make ten to a hundred times more noise than your fans? ;) You'll never worry about your PC's noise level again. Ha ha ha ha ha ha ha ha ha!

    I do basically the same thing with my car :smile:

    <font color=red>God</font color=red> <font color=white>bless</font color=white> <font color=blue>the</font color=blue> <font color=red>U</font color=red><font color=white>S</font color=white><font color=blue>A</font color=blue>
  19. the main reason Apple is still around is big ol Microsoft. If apple goes out of buisness, MS has a monopoly over the OS of the US (according to the goverment, linux is not popular enough to be in the running...yet) Anyway MS has been keeping apple alive for a few years now, I suspect they will dissapear when Linux becomes easier to use.

    "What kind of idiot are you?"
    "I don't know, what kinds are there?"
  20. it's pretty dumb really. The benchmark.

    They show one software ok. They compare the DUAL g5 1.25ghz against a single Pentium 4 2.53ghz.

    it would be like saying that a dual Athlon MP 2000+ is faster than a pentium 4 2.53ghz.

    They are compareing a dual system against a single processor system. I also beleive the dual G4 had double the RAM too!

    also it's on the Apple site so of course they are going to elaborate a lot!

    I'd take that benchmark with a grain of salt.

    <A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
  21. Quote:
    I mean how sad is that they're comparing a cheap parts, mass production setup to their top of the line one?

    They use reference style, no frills mobo/gfx boards, and I believe Lite-On drives. They do perform a little slower than same spec'd DIY'er built machines.......but not much.

    This sig runs too hot.
  22. I got sound system down too, but when using it... wife fly off the hook and b17ch at me too much. So it's low volume or headphones instead of ear bleeding by the Mrs.. :lol:

    <font color=red>Got a silent setup, now I can hear myself thinking.... great silence</font color=red>
  23. Quote:
    Udfyld felterne herunder for at skrive et indlæg til forumet. Det er ikke muligt at bruge HTML i dette forum. Tekstformatering er muligt, så det er tilladt at bruge tekstformatering i dine indlæg.

    THAT IS NOT DUTCH!!!!

    Though I don't know what it is exactly ... It is, that's for certain or Norwegian, or Swedisch, or Finnish, though the last one is very unlikely. Maybe it's even Danish, dunno ...

    Sorry for the jelling ... :redface:

    Greetz,
    Bikeman

    EDIT: I think it is Danish, as Copenhagen sometimes posts posts where the 'in reply to' is also a little ... awkward. And since Copenhagen is the capital of Denmark ... (And his profile shows he lives in Denmark ...)

    <i>Then again, that's just my opinion</i><P ID="edit"><FONT SIZE=-1><EM>Edited by bikeman on 08/17/02 01:51 AM.</EM></FONT></P>
  24. "which i posted on already =)"

    Sorry, I didn't see your post. I had to skim everyhting except for the thread because I have limited on-line time (I'm on dial-up, I can't tie up the phone line for long during the day).
    In addition, I am also highly skeptical of any "benchmark results" on any companies site (like AMD and Intel) because I've found them to be inaccurate. Perhaps they should stop cooking the performance results to sell stuff.

    Knowledge is the key to understanding
  25. <A HREF="http://www.hardwareanalysis.com/content/article/1465.2/" target="_new">check this out.</A> the dual 800MHz G4 to a single 1.7GHz P4. A 83% better performance than P4. that's more reasonable considering the p4 at 1.7GHz and crappy sdram sucked ass. but now, they're talking about a 90% performance over a p4 2.53GHz? my god what are they thinking! they'll never compare single to single, i suppose they'll have to do something to sell computers. in a couple years, they'll have to compare a 4 MP G5 with a single p4 or hammer.

    <font color=green> there's more to life than increasing its speed -Ghandi</font color=green>
  26. intel and amd are like rivaling brothers. they fight all the time, but it's nice to see people from both sides coming together when someone else attacks either of them. but there's recent news (isn't that redundant?) about apple possibly using intel/amd cpus in the future too. aww, a happy family. (would someone drop kick that guy playing the violin in the background already.)

    <font color=green> there's more to life than increasing its speed -Ghandi</font color=green>
  27. A completely unbiased (by Mac and PC users) thread on Photoshop performance is here:

    http://arstechnica.infopop.net/OpenTopic/page?q=Y&a=tpc&s=50009562&f=48409524&m=7760969205&p=16

    286 - 2x 2000 P4 Xeon
    285 - 2x 1533 Athlon MP
    279 - 2x 1466 Athlon XP
    260 - 2x 1000 G4 (OS 9)
    254 - 2x 1000 G4 (OS X)
    239 - 2340 P4 (overclock)
    234 - 2400 P4

    In the ONE application that's been massively optimised for Mac (and not PC), Apple's fastest computer is STILL lower than x86, in fact a single 2.53ghz P4 (not show in these tests) is as fast as a dual 1ghz G4.

    Of course, Apple don't tell you that in other benchmarks the G4 gets annihalated. Lightwave is a joke on Mac for example.
  28. The sad thing is that you can get a dual P4 Xeon or dual Athlon MP for about the same cost as a dual 1GHtz G4.

    Knowledge is the key to understanding
  29. LW has always sucked donkey balls on mac.

    now can we all gather around and have a circle-jerk? we need to commemorate this empty and pointless thread. shall we?
  30. Intel was pretty brave to show a 2GHZ P4 against a 1 GHZ P3, being only 80% better in video encoding. Most newbs or even enthusiasts would expect TWICE more performance, due to better bandwidth.

    --
    Is the opportunity to earn money by working, free?
  31. So uhhm... I didn't see any numbers at the bottom of the picture showing the scores. What is the world coming to? And tell me who (has some brains and editing knowledge) would buy that thing. I'm still lost on what you mean by "video editing". Are you talking about ripping DVD's or making textures? I'm sorta mixxed up about video editing, someone explain.
Ask a new question

Read More

CPUs Apple Dell