Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Technology (Apple) Businesses Apple Technology Hardware

Apple Issues New G5 Benchmarks 661

rocketjam writes "According to an article in The Register, Apple has issued SPEC benchmarks for the new dual G5 2GHz machines, comparing it to a two-way Dell Xeon and a 3Ghz Pentium 4 machine. The article says the G5 lagged behind the Dells in integer performance, however in 'the parallel "rate" benchmarks, which tax both of the CPUs in the test machines, the G5 edged out the Xeon 17.2 to 16.7 in the integer score and 15.7 to 11.1 in the floating point tests, suggesting Apple makes far better use of its two CPUs than the Xeon machine....the results augur well for Apple G5 performance in technical and scientific computing environments and for playing games.'"
This discussion has been archived. No new comments can be posted.

Apple Issues New G5 Benchmarks

Comments Filter:
  • by account_deleted ( 4530225 ) on Wednesday August 27, 2003 @10:36AM (#6805149)
    Comment removed based on user account deletion
    • rvsb quote (Score:3, Funny)

      by QEDog ( 610238 )

      Or as the guys from redvsblue [redvsblue.com] put it here [redvsblue.com]:

      The confusing thing about PCs is just that you go to the store, and there is just so many games... everywhere you look! While on the Mac, its just six. And you know which ones are good, 'cause you have already played them on the PC like five or six years ago.

      • Untrue (Score:5, Insightful)

        by Frobozz0 ( 247160 ) on Wednesday August 27, 2003 @12:59PM (#6806708)
        At the most, MacPlay and Aspyr lag 2 years. To be honest, if it's that long they rarely ship with all the problems the PC version did. Obviously I'd like simultaneous release, which Blizzard has always done, but you can't have everything.

        The Mac is easier to use. Now it's also as fast (or significantly FASTER) the the PC. It runs all the commercial apps you need. It can emulate proprietary in-house apps with VirtualPC. It can play all the latest games, even if they laga couple months (get a PS2, also!). It's UNIX under the hood and runs X11 for added compatability. All of this, and it's not any more expensive than comparable PC hardware.

        It's time to take an objective look at these systems if you're in the market for a new machine. Just take a look. If you don't like it, then don't buy it... but the Mac is a VERY viable platform these days. More so now than ever.

    • Re:Agur! (Score:3, Interesting)

      by johny_qst ( 623876 )
      Ummm... not that the availability of games for macs hasn't been traditionally terrible. And they have been getting better with the move to OSX, but I still have a problem with this augur conclusion.
      How is running a dual proc machine going to help software that isn't traditionally multi-threaded? The OS might have a fun time swapping which proc gets the game for the next 10 million cycles, but auguring well for performance.... what a farce.
      • Re:Agur! (Score:5, Insightful)

        by Aqua OS X ( 458522 ) on Wednesday August 27, 2003 @11:38AM (#6805880)
        Well, when you have Photoshop, GoLive, Illustrator, BBEdit, iTunes, Mail, iChat, Safari, Mozilla, RBrowser, MacCVS, etc all open the same time... second processor becomes very very handy.

        I typically never noticed the benefits of an MP box until I start doing web development or design on a single processor system. There are noticeable delays when switching between tasks... even on a fast machine.

        But, hey, Apple has been selling MP boxes for years now. There are a lot of applications and games that take advantage of SMP on OSX.

        Buying an MP system from Apple was the best thing I ever did. My Dual 450 g4 still feels like new to me (as long as I'm not playing games)... even with modern software.
        • Re:Agur! (Score:4, Insightful)

          by sakusha ( 441986 ) on Wednesday August 27, 2003 @12:55PM (#6806667)
          I have to agree, the multiple processor systems rock. I have a dual 1Ghz MDD machine, near the end of the line for the old G4 architecture (it's the last gen that can boot OS 9 natively, supposedly) and it is quite zippy. Don't underestimate the usefulness of dual processors. I use the old dead freeware gadget "Cycles" to watch both G4s in separate graphs, it's interesting seeing each processor's load. Apps like VPC monopolize one CPU for emulation, and offload OS and screen drawing tasks on the other processor. Some apps really max out everything, like Cleaner 6 or DVD Studio Pro, you can be up to about 98% CPU utilization on both processors, but the system is still responsive enough to toss it in the background and run other hefty apps, the main app will behave nicely and give up CPU cycles. My G4 CPU has improved memory bus bandwidth which made a really obvious improvement in performance doing tasks like encoding that are both I/O intensive and CPU-intensive. But the G5's memory bandwidth looks like it is at least 10x what my G4 can do.
          And there's one thing I think people haven't noticed. I looked at Shake 3 and Final Cut Pro, they use a new networked clustering controller called QMaster. It is a new background system service for rendering video out of Shake or FCP Compressor. You can control a whole render farm of Macs from your workstation with QMaster. This doesn't have to be a rack of XServes, it could just be the regular macs around the office. I think Apple's moving to a more networked, distributed processing model, this could be an incredible increase in computing power.
          • Re:Agur! (Score:3, Interesting)

            by jo_ham ( 604554 )
            The render farm aspect of Shake 3 and FCP was very intersting to us - we use a Dual 450 G4 as the primary editing machine (backed up with a 667Mhz Powerbook).

            Handing off renders to multiple machines would be a big benefit to us, even if it's only to make the Powermac and Powerbook work together - three processors has to be better than two in this sort of task.

            If it really works well, for a small outlay it would be pretty easy to put together a little farm of old G4 boxes - dual 450s and 500s to use them p
      • Re:Agur! (Score:5, Interesting)

        by shotfeel ( 235240 ) on Wednesday August 27, 2003 @11:44AM (#6805963)
        How is running a dual proc machine going to help software that isn't traditionally multi-threaded?

        By allowing more than one program at a time to run well.

        For example, I might have iMovie capturing DV to an external FW drive while I also doing something like play one of the games that are available for the Mac. I've actually tried and been unable to get iMovie drop out on a dual 1 GHz Mac. No matter what I throw at it, including Finder copies to/from the same external FW drive iMovie is captuing to or firing up Win XP in VPC, iMovie keeps capturing that video without dropping a frame.

        Try that with a single processor machine!
  • Apple's results... (Score:4, Informative)

    by Anonymous Coward on Wednesday August 27, 2003 @10:37AM (#6805166)
    Here. [apple.com]
  • Shocking! (Score:5, Funny)

    by mr.henry ( 618818 ) * on Wednesday August 27, 2003 @10:38AM (#6805182) Journal
    According to Apple, in certain benchmarks, the G5 is faster than the NEC Earth Simulator [jamstec.go.jp].
    • Re:Shocking! (Score:5, Interesting)

      by John Allsup ( 987 ) <slashdot&chalisque,net> on Wednesday August 27, 2003 @10:55AM (#6805390) Homepage Journal
      Well... yes... of course. This will obviously happen when the computation is to dependant on results of earlier computations that the parallelism and interconnect speed of Earth Simulator cannot help at all! I expect you can do similar with a P4 (though almost all such examples are probably heavily contrived.)

      (Recall that the clock speed of an individual processor in the ES is 500Mhz, and each processor has only 1 scalar unit, the rest being specialised vector units.) Thus you could probably contrive an example where the raw clock speed of a P4 could get the work done more quickly. I can't think of such an example myself, but someone out there probably can. ...I may be wrong.
    • Granted, the Earth Simulator has more raw processing power, but when it comes to Quake 3 the G5 has it cold.
  • by angst7 ( 62954 ) on Wednesday August 27, 2003 @10:38AM (#6805185) Homepage
    Far more interested in the progress and development of AMD's Opteron line than all this G5 stuff? I mean, dont misunderstand. I'm excited about desktop 64 bit computing, but I really dont want to be locked into a whole platform. These benchmarks really say to me that the G5 is ok, a little better, but you've gotta go all apple to get it.

    Just my pennies.

    ---
    Jedimom.com [jedimom.com], go banana!
    • by Auckerman ( 223266 ) on Wednesday August 27, 2003 @10:42AM (#6805245)
      "I really dont want to be locked into a whole platform.

      You ARE locked into a platform: x86 based computers.
    • Buy a Pegasos [pegasosppc.com] G5 when they come out and run Linux on it and you'll have yourself a quiet and energy efficient 64 bit desktop with no lock-in effect.
    • by leandrod ( 17766 ) <.l. .at. .dutras.org.> on Wednesday August 27, 2003 @10:49AM (#6805331) Homepage Journal
      >
      I really dont want to be locked into a whole platform.

      So go G5. There are two vendors of compatible processors, IBM and Motorola, while the only vendor of x86-64 is AMD, and the only vendor of IPF is Intel... not only that, the PowerPC is more efficient and has a technically brighter future.

      • by lamz ( 60321 ) * on Wednesday August 27, 2003 @10:55AM (#6805394) Homepage Journal
        So go G5. There are two vendors of compatible processors, IBM and Motorola, while the only vendor of x86-64 is AMD, and the only vendor of IPF is Intel... not only that, the PowerPC is more efficient and has a technically brighter future.

        Plus, it's what all the cool kids use. You want to be cool, right?
        • >
          it's what all the cool kids use

          I am not a kid, I have a kid.

          >
          it's what all the cool kids use

          Actually I want to have a cool, silent, energy-efficient machine. While current Apples aren't as silent and energy-efficient as I'd like, I can get a silent, energy-efficient EyeTech or Genesi.

    • by Dav3K ( 618318 ) on Wednesday August 27, 2003 @10:50AM (#6805335)
      Actually, by purchasing a G5 you aren't locked in to Apple's platform. Yes, they are the only vendor for such equipment, but that will change once IBM begins to ship their version of the same processor. Even with purchasing an Apple branded PowerPC computer, you are still free to apply whichever operating system you like - there are many versions of linux (Gentoo, Yellowdog, RH,) there is Darwin (Apple's BSD clone) or OS X. Lots of choices for lots of different platforms.

      That's not to say the Opteron isn't a cool-ass chip tho...just saying that between the two, we consumers have great non-intel options these days. I champion both companies.
      • by angst7 ( 62954 ) on Wednesday August 27, 2003 @10:56AM (#6805405) Homepage
        You make a good point (actually several people have said similar things). I guess the most important thing to me will be which chipset (G5/Opteron/Other) will have the best support among motherboard and other hardware manufacturers. Traditionally the x86 family has been the hands-down leader in this regard, and if the trend continues I dont see myself migrating to another architecture anytime soon.

        When the performance only varies by a few percent, the cost and availability will totally make the decision for me.
    • >
      the G5 is ok, a little better, but you've gotta go all apple to get it.

      For now you're right, but IBM will sell them too later. IBM prices tend to be high, but also quality; for the low end, EyeTech in the UK and Genesi in Luxembourg are now selling G4 systems. They should have G5 systems just a little down the road after IBM.

    • by Otter ( 3800 ) on Wednesday August 27, 2003 @10:55AM (#6805391) Journal
      I certainly don't propose to tell you what to be excited about -- but the point of the Apple G5's isn't that they're the unbelievablest computers ever (high-performance 64-bit workstations have been around for ages) but that it's a major step forward in the world of retail consumer computers. When Opterons start showing up in that context, running Windows, then they'll be comparable stories.
    • by Anonymous Coward on Wednesday August 27, 2003 @11:03AM (#6805490)
      If you're only interested in the G5, I'm not sure I understand where the situation differs from buying a wintelCPU product. Both sets of hardware can run Linux and BSD quite nicely.

      It seems odd to me that wintel hardware makers like HP seem more desirable because they don't make an OS to support their products. They even use software driven hardware (e.g. winmodems) that will be non-functional if you ditch the OS they bundled.

      Admittedly Macs aren't everyone's cup of tea, but finding fault with Apple because they actively develop their own operating systems (one of which is open source) is a bit like finding fault with your parents for actually raising you instead of putting you up for adoption when you were born.
    • by leandrod ( 17766 ) <.l. .at. .dutras.org.> on Wednesday August 27, 2003 @11:04AM (#6805499) Homepage Journal

      I should think before answering... third time.

      >
      a whole platform

      What does this mean?

      AFAH/WG, the PowerPC is even more open than x86-64 or IPF, having two chip vendors (IBM and Moto) while the latter have one each (AMD and Intel) only.

      AFAS/WG, the same holds true: the PowerPC runs GNU/Linux and BSD just fine, besides Apple's semi-proprietary Mac OS X instead of MS's completely proprietary Windows; the Hurd is being ported, there is Amiga-derived MorphOS and the AmigaOS itself.

      Some PCI stuff is more expensive due to the presence of x86 assembly code in firmware of dirty cheap adaptors, but this should be fixed by AMD and Intel adopting OpenFirmware's Forth dialect. So all in all, it is a more open platform.

      And BTW, if someday we get to use processors made from GNU GPL'd designs, RISC in general is a better candidate than x86-64 or IPF, being much simpler and more efficient. See that China is trying to go MIPS... MIPS in the East and PowerPC in the West would make for a much more open environment than, say, IPF in the US, x86-64 in Europe, MIPS in China and PowerPC in the high-end...

  • by BoyHowdyAAF ( 659522 ) on Wednesday August 27, 2003 @10:38AM (#6805186)

    Yes, but when are they gonna test it against the other noble gases?

    Five bucks says Argon wipes the floor with the G5, :)

  • I don't know (Score:5, Insightful)

    by cK-Gunslinger ( 443452 ) on Wednesday August 27, 2003 @10:38AM (#6805187) Journal
    I don't know if you can make the argument that "makes better use of dual CPUs" translates to "better performance at playing games." The few Opteron benchmarks I've seen have shown that it makes *much* better use of multiple CPU than the Xeons, but still lags behind single CPU system for game playing, due to the fact that dual CPUs provide little to no benefit in current games, and the SMP overhead actually reduces perfromance.
    • It's the floating point results that point to decent gaming performance in these benchmarks, and it really does seem impressive. Of course the Slashdot summary distorts the picture slightly, but that's to be expected.

      I agree that dual CPUs provide little benefit to games, perhaps if Apple standardised on two processors developers might take advantage of them?

      • I agree that dual CPUs provide little benefit to games, perhaps if Apple standardised on two processors developers might take advantage of them?

        It's not that developers don't want to take advantage of them, it's that it takes a very large amount of work to get a very moderate boost (or any boost at all, initial work on using dual CPUs for Quake 2/3 slowed the game down) from most games. Oh, and then there's the fact that most games get most of their measured performance from the video card's capabilities,
        • The Q3 benchmarks Apple posted for the 3GHz P4 don't match up with benchmarks posted elsewhere. In fact, the first Q3 benchmark I found of a P4 using a Radeon 9800 Pro was at 333 fps, 4 less than the G5 benchmark (as opposed to the 275 posted by Apple for the P4).

          The P4's Q3 benches should be much higher than these. Typical results [anandtech.com] for the P4 show scores higher than 400 fps for the fastest processors, and the linked scores are from a testbed containing an older GPU (i.e. 9700 Pro). Apple's "benchmarks" on t

      • The floating point performance of the G5 is nothing to write home about. They only managed a score of 840 in SpecFP 2000, which is WAY bellow the performance of current x86 chips (as benchmarked by anyone OTHER than Apple/Apple paid testers).

        Here are some official SpecFP (baes) scores:

        The P4 3.2GHz : 1252
        AMD Opteron 246 (2.0GHz) : 1209
        IBM Power4+ 1.7GHz : 1598
        HP/Intel Itanium2 1.5GHz : 2119

        As you can see, Apple's score of 840 in this benchmark isn't exactly impressive. Even one of the
    • Re:I don't know (Score:5, Interesting)

      by Have Blue ( 616 ) on Wednesday August 27, 2003 @11:02AM (#6805485) Homepage
      Dual-processor macs have been shipping in one form or another for at least 3 years, and a number of game ports do explicitly take advantage of them. Also, the OS is smart enough to move things like sound and asynchronous OpenGL operations to the second processor for an automatic speedup.
    • Re:I don't know (Score:4, Insightful)

      by rsborg ( 111459 ) on Wednesday August 27, 2003 @11:52AM (#6806055) Homepage
      The few Opteron benchmarks I've seen have shown that it makes *much* better use of multiple CPU than the Xeons, but still lags behind single CPU system for game playing, due to the fact that dual CPUs provide little to no benefit in current games, and the SMP overhead actually reduces perfromance.

      Well, if you're still running Windows ME, then no amount of extra processors will help your game. And, yes... games need to be multithreaded/multiprocessed in order to get any benefits from dual-CPUs (other than if you happen to run background processes in the meanwhile, at the same time).

  • by GeckoFood ( 585211 ) <.geckofood. .at. .gmail.com.> on Wednesday August 27, 2003 @10:38AM (#6805189) Journal
    Their benchmarks won't mean a thing if there's a shortage of titles for the platform and everyone buys a PC anyway. I'd love to have one of these machines, I am sure I could find some cool things to do with it. But for the price of admission, there's not enough titles out there to make spending the extra $$$ on the hardware worth it.
    • by protohiro1 ( 590732 ) on Wednesday August 27, 2003 @11:01AM (#6805463) Homepage Journal
      Well it runs everything I need, photoshop, maya, shake, Illustrator...I'm going to assume it also runs those apps very, very fast. Which is what I want.
    • by Anonymous Coward on Wednesday August 27, 2003 @11:04AM (#6805515)
      That's if you buy your computer based on games.

      Personally, I don't even like 98% of the games released for the PC. They're mostly shitty first person shooters anyway. Instead, I got a gamecube. This provides just about as much games support as i could concievably want.

      It all depends on what you want to do with your computer. There was a point where I might have considered switching from OS X to Linux, but the fact is I liked the OS X Cocoa programming environment to the point that it was enough to keep me on the mac despite the fact i hated Aqua. If you like PC games enough you're willing to let that rule everythign else about how you use your power, okay, go for it.

      -- super ugly ultraman
    • by MidKnight ( 19766 ) on Wednesday August 27, 2003 @11:11AM (#6805581)
      Are you talking about gaming titles? That is probably still the case, yes. But for any other type of software, this is just a FUD argument. In terms of being an everyday computer for the masses, it works excellently. In terms of being an everyday computer for a Java developer (my case), it beats the pants off of any other platform I've worked with.

      No, I can't play the newest PC game title, but I get plenty of gaming in on the PS2. And as a bonus, I get the warm-n-fuzzy moral feeling of not giving WinTel any of my hard-earned $$$ :)

      --Mid
  • yawn (Score:3, Insightful)

    by virtual_mps ( 62997 ) on Wednesday August 27, 2003 @10:39AM (#6805199)
    Another apple benchmark that shows intel machines running strangely slower then everybody else's benchmarks, with even fewer details then the last time we read this story. Wake me up when there's a real independent review of the state-of-the art on both systems. I wouldn't mind seeing an opteron in the mix also.
    • by Anonymous Coward
      It's easy to compare a current PC with one that's not shipping yet (the dual G5) and get favorible results.


      Gee, I could compare the latest, fastest Dell with one that will ship 6 months from now and the one that will ship 6 months from now will win!

    • Re:yawn (Score:2, Insightful)

      by dema ( 103780 )
      Another X1 benchmark that shows X2 machines running strangely slower then everybody else's benchmarks.

      Replace X1 and X2 with any variable and the statement will still be true. Thanks to the Law of Marketing. (:
    • Re:yawn (Score:3, Informative)

      by Equuleus42 ( 723 )
      Wake me up when there's a real independent review of the state-of-the art on both systems.
      According to the article [theregister.co.uk], "Apple engaged VeriTest to perform the benchmarks." That doesn't necessarily mean that the results are independent, but I would imagine that it helped.
  • Heat? (Score:3, Interesting)

    by jabberjaw ( 683624 ) on Wednesday August 27, 2003 @10:40AM (#6805203)
    Would anyone who happens to be lucky enough to own a G5 like to comment on how much heat it puts out in comparison to say, a Xenon? Just looking at the case is seems as if Apple has taken great care to make it as quiet, as well as cool, however it seems that there is a lot of space in that case.
    • Re:Heat? (Score:5, Informative)

      by Rand Race ( 110288 ) on Wednesday August 27, 2003 @12:01PM (#6806131) Homepage
      Don't own one yet, but a 400 MHz Xeon dissipates 40 watts or so and the latest 3 GHz plus models poke above 100 watts, a 3.2 GHz P4 dissapates around 85 watts, while a PPC970 at 1.6 GHz dissipates 40 or so watts while the dual 2GHz racks up about 95 watts alltogether.

  • SMP gaming (Score:4, Interesting)

    by mcgroarty ( 633843 ) <[moc.liamg] [ta] [ytraorgcm.nairb]> on Wednesday August 27, 2003 @10:40AM (#6805211) Homepage
    the results augur well for Apple G5 performance in technical and scientific computing environments and for playing games.

    On the PC, very very few games take advantage of SMP. DirectX itself seems to make zero effort, and games seem to be starting the draw from the same thread that runs the rest of the game logic. At best, you benefit a little (almost immeasurably) on I/O handling or some of the audio processing.

    Since SMP is more pervasive on Mac than on PC, do Mac games take more advantage of SMP? Does GL on the Mac render retained mode data outside of the calling thread or otherwise significantly distribute game-related work in the OS itself?

    • what? : )
    • by cK-Gunslinger ( 443452 ) on Wednesday August 27, 2003 @10:51AM (#6805343) Journal
      I agree. I would hope that with the advent (and hopefully proliferation) of 64-bit and SMP desktops, the extra processing power can be devoted to better AI. I would *love* to play against a computer opponent that requires an entire CPU to plan its moves and strategy. Get to work, software writers! Oh wait, that's me.
    • Re:SMP gaming (Score:4, Interesting)

      by Krach42 ( 227798 ) on Wednesday August 27, 2003 @10:54AM (#6805377) Homepage Journal
      PCs, and actually, more specifically Windows suck at SMP. I have a dual Pentium3 800MHz system, and Windows insists on splitting as much time as possible between the two CPUs. Thus, even if I have one and only one thread running (say, Anarchy Online) it still splits it onto two processors.

      This achieves no speed advantage, and anyone who's taken any class talking about caches, would understand why it'd be generally good to leave the task on a single CPU for at least a second or two.

      In Linux, my computer performs tons better, never "locks up" waiting on I/O. (Which is stupid of Windows, because I have two CPUs) And tasks generally split processors, but only occationally... as in, I can watch them wander back and forth.

      I've still yet to compile my multi-threaded raytracer for Windows, so I personally can't compare the speed one way or the other for real.
    • Re:SMP gaming (Score:5, Informative)

      by Daniel_Staal ( 609844 ) <DStaal@usa.net> on Wednesday August 27, 2003 @11:19AM (#6805663)
      Since SMP is more pervasive on Mac than on PC, do Mac games take more advantage of SMP? Does GL on the Mac render retained mode data outside of the calling thread or otherwise significantly distribute game-related work in the OS itself?

      Most Mac games are not specifically written to take advantage of SMP. However, OS X (which is required for a G5) is pervasively multithreaded, and distributes the load among multiple processors very well. Any thread can run on any processor, as needed.

      So, if a game is multithreaded it will use both processors. The graphics system under OS X is multithreaded, so it can use both processors. (And for that matter, the graphics card as well; Quartz Extreme offloads quite a bit of processing that way.) Basically any system call is likely to be done in a separate thread, and two threads should never take running time from each other.

      As a real-word example EV Nova [ambrosiasw.com] (one of my favorite games, so I'm plugging them.) runs much faster on my dual 867MHz MDD Mac than my uncle's 1GHz iMac, without being 'designed for' SMP. (I wish I had a real benchmark for you though.) The OS takes care of that.

    • Re:SMP gaming (Score:3, Informative)

      by mbbac ( 568880 )
      Yes, some Mac games are SMP enabled.
  • WTF? Um, old?! (Score:3, Informative)

    by DAQ42 ( 210845 ) on Wednesday August 27, 2003 @10:41AM (#6805224)
    These are the same numbers that have been up on Apple's G5 site for how long now? Since June or something? What are you people? Blind? Or just lazy. Wake me up when you get with the present. People have been arguing the validity and what not of these SPEC scores represent for months now.
  • Yes, games would be good on such a box if the vast majority of games were released to run on said box. With 90+% of the people running PC's, PC's will be where the games are found.
  • Time magazine ad (Score:3, Interesting)

    by civilengineer ( 669209 ) on Wednesday August 27, 2003 @10:42AM (#6805249) Homepage Journal
    In 'Time' magazine they have an ad every week on page 2, and this week it says the numbers are 16.9 against 16.7 for integer calculations and 15.8 against 11.1 for floating point calculations compared to a dual 3.06 GHz Xeon.
  • by rnd() ( 118781 ) on Wednesday August 27, 2003 @10:43AM (#6805265) Homepage
    Apple has issued a new release of the "Fastest Personal Computer" advertisement. The ad now reads "#1 Personal Computer Alphabetically". In a statement to the press, Apple CEO Steve Jobbs said, "Our engineers are considering alternative spellings such as Aaple in order to insure that we maintain this exciting edge in the Personal Computer marketplace."
  • OS Bias (Score:3, Insightful)

    by dlosey ( 688472 ) on Wednesday August 27, 2003 @10:43AM (#6805266)
    Could it be that the the operating system plays a role in the results? With data that close, can it really be conclusively said that the Apple hardware is faster?

    Windows never really has been that efficient in a dual processor situation.

  • Games. (Score:5, Insightful)

    by peterpi ( 585134 ) on Wednesday August 27, 2003 @10:45AM (#6805284)
    "the results augur well for Apple G5 performance in technical and scientific computing environments and for playing games."

    Well, a fast CPU certainly doesn't do any harm, but a lot of games these days are bound by either processing geometry on the GPU or by memory bandwidth for texture lookups.

    Few games are multithreaded, so having two processors isn't such an advantage.

    Still, I wouldn't turn one down.

    • Re:Games. (Score:3, Informative)

      by frankie ( 91710 )
      games these days are bound by either processing geometry on the GPU or by memory bandwidth

      Well, the (upper-end) G5 has about as much GPU [apple.com] and bandwidth [apple.com] as you can buy in the consumer market.

  • by phunhippy ( 86447 ) * <zavoid@ g m ail.com> on Wednesday August 27, 2003 @10:49AM (#6805322) Journal
    "the results augur well for Apple G5 performance in technical and scientific computing environments and for playing games."

    Right.. because of these tests every pc game developer is now going to make a port to MAC OS of all the games they are developing.. :)
  • I can see that Apple wants to emphasize that this is one really fast PC, trying to prove it by using 'un-cheatable' numbers as opposed to marketing speech, but I really think they'd look even better if they'd do some comparisions that show what a bandwidth monster those G5s are, which I have been informed is the key to how those Xeons were crushed in the showdown at MW a while ago.
  • topic Icon (Score:3, Insightful)

    by twoslice ( 457793 ) on Wednesday August 27, 2003 @10:50AM (#6805336)
    The topic Icon depicts a G4 when the topic is clearly about the G5.
  • by sniggly ( 216454 ) on Wednesday August 27, 2003 @10:56AM (#6805407) Journal
    But when will we see a 64 bit SMP version of quake3? Or any other game? Will game dev studios be required to develop multiple versions of Mac games now?

    Mainstreat Intel based OS will see the same problems I guess, 32 and 64 bit versions, linux, windows versions...

    But I bet you if id software releases doom3 for a dual g2 64 bits mac... it'll be fabulous performance compared to intel architectures.

  • by miradu2000 ( 196048 ) * on Wednesday August 27, 2003 @10:59AM (#6805440) Homepage
    Chasomint has here [chaosmint.com] a table comparing the 1.6 Ghz G5 (slowest available) to 6 other windows machines. It is a complicated photoshop benchmark. The 1.6 Ghz G5 gets beat by the single P4 3.06, however it is the 2nd fastest machine there by benchmark wins. Note that the 1.6 Ghz machine is the lowend G5, and has nowhere near the performance of the dual 2.0 ghz G5's that apple uses to test.
  • by mzs ( 595629 ) on Wednesday August 27, 2003 @11:03AM (#6805494)
    This looks to be more of VeriTest again. If you look at the Third Quarter 2003 SPEC CPU2000 Results [specbench.org] you will see that the Dells have CINT2000 scores in at least the 1100's. Apple [apple.com] gives the results for the Xeon and P4 as 836 and 889. While the Reg gives scores of 836 to the Pentium and 839 to the Xeon. The Reg article does not cite the source for the numbers in the article. I could not find anything I thought was new at Apple about this. Is this from the same statement produced by Apple with the VeriTest results from a while back? Maybe this was the first benchmark with the dual 2Ghz G5's?

    In any case there was much consternation in the past about the VeriTest benchmarks becuase they did not use the same compilers that Dell used. Also VeriTest used things like an optimized malloc library on the G5's and faster memory with semi-secret memory timing tweaks in OF. If you want to take these benchmarks with a grain of salt, you should compare the DELL numbers from the SPEC site to those of the G5 from Veritest.

    • by LemonYellow ( 244336 ) on Wednesday August 27, 2003 @11:10AM (#6805574)
      I think we've been through this before.

      If you want to test relative compiler technology, you use the fastest compiler for each platform. If you want to test the platform itself, you use software which is as close to identical between platforms as possible. Hence, gcc.

      Pretty basic experimental technique is at work here.
      • by janolder ( 536297 ) on Wednesday August 27, 2003 @12:44PM (#6806547) Homepage
        What you're proposing as basic scientific technique is bad science. Gcc's backend is quite different for G5 and x86. Testing this way invalidates the results. It's almost like testing both a diesel truck and a Ferrari with diesel fuel on the pretense of fairness. The truck wins hands down, but have you really learned anything?

        If you want to compare compilers, you run different compilers on the same hardware. If you want to know how fast the hardware is, you let the manufacturer hand-tweak the test as much as possible, which includes picking the compiler and its options.

        If you want to know how fast the hardware is in the eyes of the user, you use standard configurations which the software vendors will use: GCC for G5 and MSVC or Intel for x86. Using an inferior compiler on x86 (the x86 backend of gcc isn't that great, in fact it is quite bad) doesn't mean squat to the user. Word/Excel/Photoshop isn't going to get compiled on gcc for x86 anytime soon.

    • You'll also notice that Dell uses special tweaks and memory management libraries to get their results.

      However, the nature of SPEC is that you perform your own tests and submit them. You certainly should not be performing any other manufacturers tests for them...

  • by silverhalide ( 584408 ) on Wednesday August 27, 2003 @11:06AM (#6805529)
    The only benchmarks that should matter are: does it run Photoshop/Premiere/Final Cut Pro faster? I could care less that my computer is a 11.5 compared to 10.32 on another machine. That means nothing to the end user.
  • by PghFox ( 453313 ) <afoxson@nospAm.pobox.com> on Wednesday August 27, 2003 @11:20AM (#6805679) Homepage
    The Apple Developer Connection Student Program [apple.com] is a low-cost membership program (USD $99 / year), providing tools [apple.com] and special discounts [apple.com] for students 18 or over interested in developing for the Mac platform. Members receieve a once-per-lifetime 20% discount on hardware [apple.com]. Hardware can be purchased through the ADC version of the Apple Store [apple.com] (click the 'ADC Hardware Purchase Program Store' link). Without the discount a Dual 2 GHz G5 would be USD $2999, and with the discount a Dual 2 GHz G5 would be USD $2499. Details of the program are covered in the FAQ [apple.com].
  • by alchemist68 ( 550641 ) on Wednesday August 27, 2003 @11:34AM (#6805829)
    Apple's Marketing department needs to show television commercials showing how EASY it is to network their computers to existing Windows and Linux corporate networks and continue running in the event of a virus/worm breakout in the Wintel world. Apple needs to show some compatibility with its computers if they ever intend on selling more units to new customers. Apple needs to show that its computers offer LOWER COST OF OWNERSHIP for many computing environments. I simply don't understand why Apple can't grow a spine and show a hint of agressive advertising. Apple has a really good product, both hardware and software, but people aren't going to buy their computers unless they know they aren't going to be stranded ALONE in the computing world when a slight hiccup occurs with hardware or software. One of my friends who works in business management put it like this: "I'm worried about support for software and hardware. Nobody uses Apples in business because they're not supported". While I disagree partially with my friend's statement, there is a little truth to it. If Apple's in the computer business to make money, it better start showing ADVANTAGES of its products compared to the DISADVANTAGES of competition, otherwise no one will notice their product offerings. With The Borg's long history of patches for its products not doing a very good job of protecting the security of networks, email, documents, etc... Apple could begin with an advertisement detailing those advantages.
    • Apple's Marketing department needs to show television commercials showing how EASY it is to network their computers to existing Windows and Linux corporate networks and continue running in the event of a virus/worm breakout in the Wintel world.

      Except that SMB services on Jaguar kinda suck. There are a number of bugs. WINS name resolution, for instance, doesn't work. So if you use NT Domain services (not Active Directory) you can't browse shares on a routed network. Bit of a problem for many corporate user

  • by Nom du Keyboard ( 633989 ) on Wednesday August 27, 2003 @11:41AM (#6805929)
    This is all marketing. Apple is making comparisons against P4 and Xeon computers instead of the not yet shipping Athlon-64 systems to make people feel good about buying Apples. Too many Apple owners are belittled by their PC (and I don't mean Politically Correct) friends for having underpowered machines. These ads are intended mostly to give Apple owners ammunition to fight back now. The top-end desktop machines are on an even footing again, and that's the reputation Apple needs to re-establish.

    Marketing, pure and simple -- and effective.

  • by Transcendent ( 204992 ) on Wednesday August 27, 2003 @01:10PM (#6806794)
    the G5 edged out the Xeon 17.2 to 16.7 in the integer score and 15.7 to 11.1 in the floating point tests, suggesting Apple makes far better use of its two CPUs than the Xeon machine

    Edging out now implies that it's far better?
  • by raph ( 3148 ) on Wednesday August 27, 2003 @03:43PM (#6808137) Homepage
    If you look at the numbers Apple is claiming vs. the latest numbers posted at the SPEC site [specbench.org], then it would appear that the G5 is getting creamed by the Pentium 4 / Xeon.

    First, single CPU performance. Apple claims 840 for SPECfp_base2000, and 800 for SPECint_base2000. A Dell Precision 360 with 3.2GHz P4 and DDR400 memory gives 1267 and 1242, respectively.

    Next, dual-CPU. Apple claims 15.7 for SPECfp_rate_base2000 and 17.2 for SPECint_rate_base2000. A Dell Precision Workstation 650 with dual 3.06 GHz Xeons gives 18.0 and 25.6, respectively.

    Of course, there are lies, damn lies, and benchmarks, but in this case I think it's fair to compare actual SPEC numbers with vendor claims.

    And don't get me wrong, I think Apples are wonderful systems. I recommend them to many of my friends. But for raw CPU power, they lag the Intel powerhouse.

The bomb will never go off. I speak as an expert in explosives. -- Admiral William Leahy, U.S. Atomic Bomb Project

Working...