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

 


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

Yet Another G5 Roundup 69

Posted by pudge from the you-thought-we-were-done dept.
Lawrence Person writes "This article on Low End Mac talks about why the PowerPC 970 is so fast, covering its superiority to Intel chips in Multiply Accumulate, double precision arithmetic, and Fast Fourier Transforms, among other operations. A short, clear article for those who don't have the time to wade through Parts 1 and 2 of Ars Technica's exceptionally detailed dissection of the 970/G5." Trollaxor writes "IBM has a neat two-page history of the PowerPC architecture, detailing its evolution from the first RS/6000 chipsets in 1990, through the POWER ISA, and into the processors that we know and use today. A very interesting read."
This discussion has been archived. No new comments can be posted.

Yet Another G5 Roundup More

Yet Another G5 Roundup

Comments Filter:

  • FFT is a good mesure (Score:5, Informative)

    by Baikala ( 564096 ) on Monday July 07, 2003 @12:59PM (#6383718) Journal
    Fast Fourier Transform is bread and butter for the scientific comunity. This is a good news for sys admins at research centers like me.
    Maybe I have a chance at getting one or two of these babies for the next year budget.

  • Okay (Score:5, Insightful)

    by Japer Lamar Crabb ( 670674 ) on Monday July 07, 2003 @01:15PM (#6383822)
    I understand the excitement over these machines, so I won't get all pissy about this, but...

    Until these machines are widely available, each and every thread concerning the performance of the PPC 970 will run the risk of degenerating into a heated debate over whether the figures being offered are reliable. In other words, a flamefest.

    Don't we already have enough of those around here?

    • Re:Okay (Score:5, Insightful)

      by Johnny Mnemonic ( 176043 ) <mdinsmore@NoSPaM.gmail.com> on Monday July 07, 2003 @02:15PM (#6384275) Homepage Journal

      Hey, I'm a Mac fan, and I totally agree--while this bodes well for the future, the benefits of the G5 are clearly application specific. So until you'll be able to test it yourself, with the apps that you use, compiled with whatever they're compiled with, it's just another flamefest.

      That said--both you and I clicked the link to "Read More", didn't we?
    • they get a damn OS that is built specifically for a 64bit processor instead of a 32bit patched. It's kind of like running DOS on a Pentium, pointless.
      • Hey, as soon as I can buy a PC without a A20-gate to allow just that (running DOS on a Pentium - or anything from a 286 infact), I may buy one.

      • Re:I just think it's pointless until (Score:5, Interesting)

        by dhovis ( 303725 ) * on Monday July 07, 2003 @03:52PM (#6385127)

        Actually, that isn't really true. Because of the way the PPC ISA is set up, there isn't much advantage to switching the whole OS to pure 64 bit. The main boost you are going to see here is the ability to use more than 4GB memory, and you can implement that on the G5 with just a recompile.

        Individual apps can be switched to 64-bit and operate in 64-bit mode if needed, even if the OS doesn't use 64-bit mode itself, so long as the OS supports 64-bit addressing

        Unless you need to do 64bit integer math, your app will see no benefit from switching to pure 64-bit. In fact, your app may slow down and waste memory. But an app can still take advantage of the main benefit of more addressible memory whilest staying (essentially) a 32-bit app. This is thanks to the fact that PPC is a 64-bit architecture with a 32-bit subset.

        There are still optimizations that need to be done to improve the compilers for the G5, but very few of those optimizations have anything to do with being able to do 64-bit integer math.

        • Let's face it, the most relevant and useful thing we're likely to get from 64-bit desktop applications and systems is going to be the ability to keep time from ending in 2038.

    • whether the figures being offered are reliable (Score:4, Interesting)

      by dpilot ( 134227 ) on Monday July 07, 2003 @02:40PM (#6384458) Homepage Journal
      And of course they're being compared with P4 numbers that are now "mainstream." But when the P4 was first introduced, it was "peaky" and irregular, behaving much different from the well-understood PIII and K7 cores. AFAIK, aside from speed bumps, both internal and frontside, and cache size increases, it's still essentially the same "net-burst" core that received such mixed reviews on introduction. Oh, and quite a bit of compiler work, I'll guess, not to mention the new SysEnter stuff under Linux.

      Intel got much-deserved heat on the P4 introduction, though that seems forgotten now. IMHO the early irregular performance seems to have been handled by tweaking compilers and ramping speed until the valleys are mountain glens. For that matter, Merced seems largely forgotten with McKinley and Madison. Adoption has simply happened over time, because it's Intel.

      But there seems to be an air about that everyone else's (PPC970, K8) difficult launch is nearly fatal, and we should wait to adopt until these issues are ironed out. Of course many of them are volume-related and won't be fixed by anything but production and experience, same as P4 and I2.

      We seem to be a bunch of monopoly-making sheep, more times than just this one.

  • Yea, but does it run Linux? (Score:5, Interesting)

    by Znonymous Coward ( 615009 ) on Monday July 07, 2003 @01:21PM (#6383864) Journal
    Seriously. The G5 PowerMac has like 9 fans [apple.com] in it that are controlled by the OS (Mac OS X). It will be easy to run Linux on it, but will Linux properly control the fans to keep the system from burning up or flying off the desk?

    • Re:Yea, but does it run Linux? (Score:5, Informative)

      by dhovis ( 303725 ) * on Monday July 07, 2003 @01:47PM (#6384023)

      I think the tech documents say that if the OS does not provide thermal management, the fans will run at full speed. So it will be interesting to see if the thermal management will be provided under Linux.

      Terrasoft [terrasoftsolutions.com] (makers of Yellow Dog Linux) has said that they will support Linux on the G5, but it remains to be seen if they will be able to provide thermal management that won't void the warrenty. Terrasoft is an offical Apple Value Added Reseller, and they sell dual boot MacOS X/Linux systems that carry the full Apple warranty, so Apple may provide them with the info they need or else a binary driver that they can use.

      It may also be that Apple will make the thermal management code open source as part of Darwin. If that is the case, then it can probably be converted into a kernel module without violating either the GPL or APSL.

    • Re:Yea, but does it run Linux? (Score:4, Interesting)

      by Duke Thomas ( 684070 ) on Monday July 07, 2003 @01:51PM (#6384056)
      I never thought of that. Why did Apple put the OS itself in charge of regulating the temperature? Why couldn't the fans be controlled by some simple thermostat like mechanism quite apart from the operating system? Apple has been big on software controlling hardware for a while now, but this seems a bit much.
      • This may be a logical answer: its much easier for them to update the code running the fans from the OS than in firmware.
      • I think there are several factors:

        1. Like you said, ease of updating the code
        2. The OS can predict when the CPU, hard-drive, video card will get hotter since it knows when they're being used, and thus can turn up the fans in those areas a little before the use spikes, whereas sensors on the fans can only react to an already existing temperature increase - not as good
        3. Anything that forstalls Linux deployment on Apple PPC will mean increased sales of add-on software, including third-party software. This ec

    • Comment removed (Score:4, Funny)

      by account_deleted ( 4530225 ) on Monday July 07, 2003 @02:19PM (#6384305)
      Comment removed based on user account deletion
      • The design is really to minimize noise I think because if you have lots of slow moving fans, it's a lot quieter then one or two fast moving ones.

        Computers put out too much heat to use passive cooling anymore, the old computers of yore didn't use fans, but people today don't want slow machines so we have to use active cooling.
      • Brrrruuurrrp! One of the two Serial ATA drive bays is taken by the HD, the other is free. The optical is in an extra bay.
      • If a fan fails, it's not necessarily going to kill the computer. Considering Apple says they used 9 fans to reduce noise, something tells me that the failure of a fan will not be a catastrophic failure in terms of the entire system. If the OS is controlling the speed of the fans, then the OS should be able to be written to either limit speeds or shut down the system upon the failure of a fan, so it shouldn't be a real issue. I really don't see why so many people are all hung up on the fact that it uses 9 fa

      • Re:Yea, but does it run Linux? (Score:5, Interesting)

        by MrTangent ( 652704 ) on Monday July 07, 2003 @03:00PM (#6384642)
        There are actually two hard drive bays. Look closely.

        The lack of more than two hard drive bays was a concern of mine as well, but the drive that ships with the G5's is 60 to 250 gigabyte Serial ATA. Factor in the plethora of fast Firewire/Firewire 800 external drives and most Mac users will be as happy as pigs in shit (as well as Fibre channel XServe Raid for multi-terabyte storage for the higher end users). So you can get half a terabyte internal Serial ATA in the G5's if you chose to pay the premium. I think that's more than enough for 99% of the population.

        I do agree there's some aspects of the G5 design that seem like a step backwards over the previous G3 & G4 'El Capitan' case. While two hard drive spaces is fine for most users, the G4 allowed you to have four internal hard drives. The lack of a second optical drive bay is also a "step backwards" over the G4 design (this can be rectified with an external Firewire drive if needed, and again, most users don't require a second optical drive).

        For me though the biggest gripe I have with the G5 case is the removal of the drop-down motherboard door design on the G3 (Blue and White) and G4. The one button access with drop-down door was an amazing feat of engineering and it saddens me to see this go. The new G5's internals are very elegant but installing and upgrading RAM, PCI and AGP cards is so much easier on the previous G4 and G3 models (although the G5 certainly looks easier to upgrade/get into than most PC boxes I've worked on before).

        As far as the fans go, I have no gripe there. While it might be more moving parts, the idea of thermal zones is a sound idea and will help keep the machine cool and output less decibels. By the way, the fans will be replaceable. You make it sound like the machine will be junk just because one fan fails. I'm sure Apple has thought of this and will provide replacement fans.

        One could also argue that by having nine fans that work in tandem you'll have less chance for failure since no one fan will be running at 100% at all times like on other machines. This quote kind of explains what I mean:

        "the candle that burns twice as bright burns half as long."
        • Comment removed based on user account deletion

          • You forget that Apple's thermal managment software is constantly monitoring things and if the temperature goes to high, the machine goes to sleep. In fact, the G4 Cube would sleep if you set something on top of it and blocked its ventilation. It didn't even have a fan.

            In fact, the 9 fans give you some amount of redundancy. Under normal operation, they turn at a low fraction (10%, IIRC) of their top speed to stay quiet (just 35dBA), so if one fails, the others can take up the slack with no problem. The f

      • I've used PCs now for decades, have Amigas and Sinclair QLs and, goddamnit, Dragon 32s still in working order, but I don't have a SINGLE fricking machine that relies upon a fan working that's more than three years old.

        For PCs, this may be true, but for better hardware, such as (hopefully) the PowerMac G5, the fans should last at least five years, not three. This has been my experience with Sun workstations, where the ball-bearing case fans begin crapping out after five years of 24/7 operation (five year

      • There are two things (yeah, just two) I'm not 100% happy about on the G5 series:

        1. The relative lack of expandability (there are only two drive bays, for example, and both are already in used by default - one HD, one optical media)
        2. NINE FANS! Holy crap Apple, what the HELL were you thinking?!

        OK, so others have pointed out that you get two hard drive bays, and the SuperDrive is standard, making that a bit less of an issue in my mind.

        About the 9 fans, if they turn out to do a better job of managing

      • The relative lack of expandability (there are only two drive bays, for example, and both are already in used by default - one HD, one optical media)

        Wrong. You get one optical, but there are two bays for SATA hard drives. So you have one unused bay. What I don't like is that there is no ATA backwards-compatibility, so it looks like I'll be buying an external FireWire case for the hard drive in my G4 now, if I want to keep it. I don't think you have the option of adding an ATA PCI card, either, because you'd have a tough time running the cable to the drive bay without interfering with the cooling zone setup.

        NINE FANS! Holy crap Apple, what the HELL were you thinking?! I mean, nine fans. That's NINE moving parts. And if any of them fail, presumably that's the end of a $3,000 computer.

        I'll tell you what they were thinking: Nine fans spinning slowly and quietly move as much air as the two or three fast-spinning, deafening fans in the average PC these days.

        Also, your $3000 G5 will be much better designed than the average PC. Have you seen the clear insert in behind the removable side panel that divides the cooling zones? That's not there just to look cool, it is there to properly duct the airflow to where it is needed in the mcahine. I read in one article that the computer will not even power on if THAT is not in its proper place-- so I'm feeling rather certain that Apple has taken fan failure into account, and if necessary you'll either get a warning and have a minute to do a graceful shutdown manually, or the computer will promptly, gracefully (or not) shut down on its own.

        ~Philly
      • >> The relative lack of expandability (there are only two drive bays, for example, and both are already in used by default - one HD, one optical media)

        There are 2 SATA drive bays for up to 500 GB internal storage, AND an optical Drive. Further more, you can always add external devices through USB2, Firewire 400 / 800, gigabit Ethernet.

        >> NINE FANS! Holy crap Apple, what the HELL were you thinking?! I mean, nine fans. That's NINE moving parts.

        This is a Good Thing - distributed cooling. If 1 of
      • You can use fans with integral speed sensors. This allows the system software to detect failing or dead fans. I don't know if Apple uses them, but they have been available for many years.
    • betcha a dollar they GPL a driver for the fans.

  • Ah yes, the early days! (Score:5, Interesting)

    by dbirchall ( 191839 ) on Monday July 07, 2003 @01:43PM (#6383998) Journal
    In April of 1990, I was visiting a friend at Utah State University in Logan, UT. The IBM rep (nice lady) came around to the computer center (which was nice of her) to try to sell everyone PS/2's - or at least PS/1's (which was less nice of her).

    But she brought this one thing with her that looked kind of like an overgrown PS/2, and had a goshawfulbig monitor hooked to it... and was running UNIX. Being a geek even back then, I noticed this and asked what it was and if I could play with it.

    'Twas some very early RS/6000 model, quite unstable at that point in time, OS-wise. I have no idea why she was allowed to bring it on campus. Maybe she was trying to convince them to move away from their Ultrix vaxen.

    By IBM's timeline, that would have been a POWER (no numbers after it) chip, predating the PowerPC by a chunk of time. I never stopped to think about it before, though.

  • Multiply and add (Score:5, Interesting)

    by Matthias Wiesmann ( 221411 ) on Monday July 07, 2003 @01:54PM (#6384084) Homepage Journal
    One thing that irks me in the low-end mac article is that it states that the G5 can do a multiply add in one cycle. While this is true, this is nothing special about the G5, the multiply and add instruction has been in the PowerPC instruction set since the start - my Powermac 7100 (technically à G1) already could do this. This is in fact pointed out in the intersting article by IBM about PowerPC.
    • Actually, your PM 7100 is technically a G2. G1s were 68k-based macs. G2s were PPC 60x-based, G3 is PPC 750-series processors, and of course PPC 74xx = G4, and PPC 970 series (so far) is the G5.

  • Apocalypse Now (Score:3, Funny)

    by Johnny Mnemonic ( 176043 ) <mdinsmore@NoSPaM.gmail.com> on Monday July 07, 2003 @02:11PM (#6384236) Homepage Journal

    Fast Macs, and trollaxor with a story submission. The apocalypse must be near now...
  • Wasn't the PowerPC line supposed to be a RISC chip, anyway? Why would they (IBM, not the article) feel the need to proclaim that they support one more instruction, anyway?

    I'm sure we can find particular instances where the x86 has an instruction that the PPC doesn't have... after all, the PPC is supposed to gain its speed from having fewer instructions, right?

    ahhh head is spinning from too little sleep... beat with cluestick as necessary.
    • First, RISC vs. CISC arguments died about a decade ago. What was RISC became Load/Store architectures because they have load and store operations for memory access and everything else is register based. Most so called "RISC" machines tended to have nearly as many (if not more) instructions than their "CISC" counterparts.

      I posted a while back (sometime in the past year) the number of instructions a G4 (including Altavec) has compared to the P4 (including SSE2) and the G4 had quite a few more opcodes than
    • the point is, RISC does not implies there are less instructions for you to use on the processor. What RISC actually means is that most instructions can be finished in one clock cycle (LOAD and STORE may be an exception) and all instructions are register based (as pointed out in sibling comment). Since all of its instructions are register-based, it can finish in one cycle (memory fetching takes definitely more than one cycle, compare your CPU clock to (effective) memory clock) and more importantly, it does n

  • Floating Point Units (Score:2, Informative)

    by yarbo ( 626329 )
    comparison table [amd.com] the Athlon XP has 3 full FPUs, the P4 has a full and a partial (I believe it only can do memory operations, not arithmatic).

Slashdot Top Deals

"Ninety percent of baseball is half mental." -- Yogi Berra

Close