maximumpc benmarks dual g5 vs amd vs intel

Discussion in 'Buying Tips, Advice and Discussion (archive)' started by superbovine, Nov 27, 2003.

  1. superbovine macrumors 68030

    superbovine

    Joined:
    Nov 7, 2003
    #1
    December Maximum PC has a great article on benmarks of the new dual g5 vs amd 64 bit vs intel new chip. The benmarks weere done with the help of MacAddicts Magazine for those of you who think the Max PC would cook the benmarks. The benmarks look fair, and the analyses is good, but the article is a little biased towards PC. All in all it shows that most software hasn't been optimized for the using multi processor yet, so most benmarks are not what you'd expect. i believe once that happens g5 will start kicking some booty.
     
  2. skymac macrumors member

    Joined:
    Aug 4, 2003
    Location:
    Toronto,On
    #2
    In the most recent Mac World there was also a good comparisn between the dual g5 and several P4 or zeons etc. the g5 only one in 10r 2 of the benchmarks AND one of the Pcs were 2 dual processor. the g5 still has some way to go but buy the next update they will really start to kick some even more serious butt. i just hope that apple can catch up the pc market in terms of clock speed within the year.
     
  3. i_wolf macrumors regular

    Joined:
    Jul 17, 2002
    #3
    point in fac the macworld results were arseways. Different apps being used to compare. does not a benchmark make!
    Premiere on Mac is years old and very legacy. Only runs in classic which is emulation mode and is only uni processor. They compared this with latest optimized bang up to date premiere for pc. surprise surprise, an optimized piece of software running on a dual processor wins out against a dual processor in "uni processor mode" i.e. not SMP aware, running an old legacy unoptimized app in an EMULATED operating system on top of OS X. Brilliant. just brilliant.
    MP3 encode test..... use two totally different apps with different mp3 encoders. Sure on my pc i get different results encoding with LAME (1/3 speed) compared to Music Match :S Again brilliant comparison by macworld.
    Another test they showed 4 bench's in photoshop. With a little disclaimer saying that had they taken into account loading times of app and loading times of filters mac would have won by a significant margin. Also they tested on Jaguar... but don't think it would have made much difference on Panther. They used teh G5 plugin which optimizes a few filters but is questionable what improvement its actually giving. Macworlds bench's in this regard were also totally questionable. Compare that with a well renowned website and reliable website like arstechnica where they did a full comparison using photoshop between x86 and G5. Funnily out of 30 filters G5 won over 95% of their filter tests. Incidentally the G5 plugin doesn't affect the majority of these plugins.
    Looking at their quake 3 results. What a joke. Different video cards, different memory and runnin jaguar with older video drivers. Not valid.
    Microsoft Word. I never new that Word was considered an industry standard for benchmarking latest and greatest processors :S
    All in all that review was a joke.
    To get real world indicative ideas of perf and expected perf check this .
    http://apple.slashdot.org/apple/03/08/28/1746213.shtml
    Also check the arstechnica forums. Recompile of a lot of older apps gave the G5 a double or tripple lead over the x86 equivalents in many tests by testers in the arstechnica forums. They were using a new compiler which can generate PPC970 optimal code (code that can make full use of the wide and deep pipelines and use all the cool trick the chip has up its sleeve like dual fpu, dual int etc..). Apples compilers while still very young do not do this. Have fun, but by all means do not just go by those "benchmarks" in macworld. Funnily enough, in the December issue i had expected a huge backlash in the letters section but it was strangely quiet. Says everything really.
     
  4. ddtlm macrumors 65816

    Joined:
    Aug 20, 2001
    #4
    I get the feeling that some people believe anything other than victory for G5's is "biased". So the G5 wins some things and looses in others, get over it, move on.
     
  5. hulugu macrumors 68000

    hulugu

    Joined:
    Aug 13, 2003
    Location:
    the faraway towns
    #5
    More benchmarks yay

    Seriously do these things ever really mean anything? AMD vs. PIV vs. G4 vs. G5 vs. etc. Every single time there's some strange variable that cannot be accounted for, and some bias that is always there from either side. I wish someone who didn't have a stake in any system would go through and complete a system by system with controls test. But, until that happens I will ignore any benchmark that comes out.
    The entire system makes a bigger difference than the speed of the processor—memory, FSB, BS Cache, video card, drive speed, etc. that matter on the user's side, which is really what matters anyway.
     
  6. m4rc macrumors 6502

    Joined:
    Sep 15, 2003
    Location:
    UK
    #6
    A few comments. Firstly, you guys need to take a little more time over your posting, the formatting, spelling and punctuation is so bad that they are really hard to read. Not an attack, it's a fact.

    Secondly, yes, will you settle for nothing but 'The G5 wins every test'? We KNOW it isn't the fastest out there, as ddtlm said, get over it.

    Thirdly, what does it matter? If you own a G5 and a magazine article prints a test showing a PC beating your pride and joy, does it make your G5 slower? No. It's the same as when a new, faster Apple comes out, your older slower system dosn't suddenly slow down or become a poorer system, it has just been overtaken by something faster, and as you all should know by now this will always happen..

    You all know how fast your own system runs, it's strong points and weak points. Just because someone in a magazine points out that a PC can apply a photoshop in just under 4 minutes whilst your G5 takes 4 minutes and 1 second, does it really matter?
     
  7. Dont Hurt Me macrumors 603

    Dont Hurt Me

    Joined:
    Dec 21, 2002
    Location:
    Yahooville S.C.
    #7
    The bottom line is the Mac is back with the G5, imagine if all we had were the lousy G4 machines, the G5 puts us back on a level playing field. the sooner this chipmakes it to other apple products the better. im glad we had a G5 to test
     
  8. i_wolf macrumors regular

    Joined:
    Jul 17, 2002
    #8
    for a start its a computer forum. not an english forum. get over it. Also for what its worth this is my second language. Like to see you speak irish with good semantics, grammer and spelling.

    it isn't a matter of some pc beating a pride and joy. im lucky enough to have access to Xeon, Opteron and G5. The point of the matter is that there will be people out there looking at these benchmarks in the magasine and rightly so want to get the best going. The benchmarks are very misleading. Thats my point. They are particularly misleading because they don't have similar specification machines being used. They aren't using similar builds of applications (adobe premiere). They don't tell you that they have one machine running in emulation mode in non SMP against an up to date optimized SMP aware version on pc. That isn't a benchmark. that is bad journalism. They compared different mp3 encoders... again thats bad journalism.
    My point isn't so much about them using optimized or unoptimized but that they used different apps on different platforms which wasn't fair. It could have been fair. Say they took LAME and used that as a measure of mp3 encoding performance. That would have been a good bench because the source code is available for free. They could compile it with the optimum compiler for each platform and then bench. Thats thorough.
    The premier thing was a joke.
    Only a small subset of filters were used in photoshop. Why not test all of them and give an overall which is fastest. Arstechnica, did something along those lines. G5 beat the Opteron , Xeon by a significant margin wherever it won a filter, and was pretty much neck and neck when it was behind. Incidentally the G5 according to ars beat out the Xeon and Opteron in something like 25 out of 30 filters. Chaosmint did the same and found the G5 significantly beat the opteron and xeon whenever it won in 15 out of 21 results. And was neck and neck in the rest.
    The problem is that macworld didn't do a thorough benchmark and didn't use a good methodology for testing (namely the same app, on similarly configured hardware) which isn't as difficult as some people make out.
    Now this post i found interesting. can you explain the we KNOW that the mac isn't the fastest out there. If anything it isn't running at its full potential at all at the moment. Opteron , Xeon all can they have good compiler support and their architecture doesn't really depart heavily from previous x86 generations. For example athlon 64 is based on athlon xp so established code for athlon xp runs happy on opteron and is already optimized more or less in 32 bit mode. P4 was a duck to swan in performance terms. Both P4 and Opteron can automatically take advantage of SSE2 generated by autovectorising compilers like GCC.
    GCC doesn't autovectorise its code. It is planned that this feature will make its way in. Currently IBM's XLC and XLF compiler do not do the same either. However they generate code for the ppc970 which does make use of its radically new design. Remember that the 970 is radically different to G4 and G3 so legacy code just doesn't get to run the chip at anything near its theoretical performance. Arstechnica when they disected the G5 said that presently it should be neck and neck with Opteron because it can't run at its theoretical max performance. Ars said that its normal perf on day to day GCC optimized code would be 50% its theoretical performance. Reason for this is that again its design is very different to older PPC designs. wide and deep execution lines. While it may haev dual complex fpu, dual complex int units. GCC does not generate code that allows the scheduler to fill the wide pipes. it doesn't allow the chip to run both complex int, fpu units in parallel as they were designed. and it certainly doesn't allow for the theoretical perforance of having 215 in flight instructions at once. Ars has pretty much been proven spot on since the IBM compilers have been released. they generate "compatible" code for PPC 970 (although they still don't autovectorise for alitvec it is planned). Recompiles with this code has shown two to three times the performance over what GCC currently produces. Again this is a very young chip architecure and the compilers are still very young. its likely that performance will improve again from compilers as they improve. But as a first example of what the chip can do with the correct compiler its awesome. Easily faster than opteron and Xeon. Check the thread out at arstechnica.
    http://arstechnica.com/archive/news/1062961031.html
    and then the thread where people recompiled their real world apps with this compiler.
    http://arstechnica.infopop.net/OpenTopic/page?q=Y&a=tpc&s=50009562&f=8300945231&m=4250963285&p=1
    The general concensus is that when you use all the execution units of the chip by using the IBM compiler performance (available for free presently from ibm's website) it improves performance two to three times what GCC produces.
    So if G5 is neck and neck (sometimes faster sometimes slower) on code generated by a very unoptimized GCC compiler, which doesn't really take advantage of all the bells and whistles the chip offers (extreme parallelism) and which can't autovectorise like it does for SSE2 on P4 and opteron; just imagine whats gonna happen when more and more apps get recompiled with this new compiler. And thats the point. Can Opteron and Xeon say that presently they aren't working in optimal conditions and are only running at 50% efficency with current compilers?? Can either lay claim to 200 to 300% performance improvement from a recompile (this isn't based on alitvec by the way ;) ) ???
    I think not. And thats prob the biggest point that a lot of magasines are missing.
     
  9. psycho bob macrumors 6502a

    psycho bob

    Joined:
    Oct 25, 2003
    Location:
    Leeds, England
    #9
    I agree that the mac world lab tests were a waste of space. Anyone using premiere on the mac still when FCP 4 (or even 3) are available is mad it is vastly superior. Although all the photoshop tests on the whole include adobe's 25mb G5 update I fail to see how this provides full optimization. We need more independant tests with Adobe CS now that it is available. What is holding both Apple and AMD back is the lack of a 'from the ground up' 64 bit OS. At least PC users do have a 64 bit capable version of XP arriving soon (think it's in beta at mo). Once a 64 bit OS is out specific 64 bit apps will follow.
    For the first time in many years Intel really aren't in the race. They are turning to overclocking and that good apple trick of L3 cache with their latest processors. In the PC field despite AMD's 64 bit processor being down by around 1GHz compared to Intel's latest it beats it on many benchmarks basically all except some games needing raw power and software which can utilise hyperthreading and large amounts of cache. A good website for PC news is Tom's Hardware, I believe they were the first to build a 4GHz PC that ran stabily by overclocking P4's.
     
  10. m4rc macrumors 6502

    Joined:
    Sep 15, 2003
    Location:
    UK
    #10
    My apologies if I offended you over your spelling, it was just very hard to read the first coouple of posts, no offence intended. No, I can't explain that we KNOW the G5 isn't the fastest computer, sorry. But I can say that we can't prove it is. If it was provable, Apple would not have had the plug pulled on the ad's.

    You guys are making far too much out of this, it's the PC users who are meant to get upset that we can work smarter and faster. Even if the machines run slightly slower - whether they do or not I don't actually care - it is a fact that Apple's OS enables us to work with so much more ease. I use a PC all day, and a PowerBook at other times. I dont have a figure, but am so much more efficient on the PB, and enjoly using it more.
     
  11. psycho bob macrumors 6502a

    psycho bob

    Joined:
    Oct 25, 2003
    Location:
    Leeds, England
    #11
    well said. I don't use macs for their raw performance.
     
  12. topicolo macrumors 68000

    topicolo

    Joined:
    Jun 4, 2002
    Location:
    Ottawa, ON
    #12
    Here's a thought: What does it matter if one processor is THEORETICALLY faster than another? If this processor is slower in real world tests, all of those SPEC scores wouldn't make a difference.
     
  13. i_wolf macrumors regular

    Joined:
    Jul 17, 2002
    #13
    First off, apologies that i got a bit over sensitive with the spelling. You are right, in that if you want to make a point make it easy for readers to read it. Mates :)
    you see the point isn't whether one chip can one a nano second faster than another. on present day apps i don't care because certainly for a lot of modern apps, a xeon , opteron or G5 is extreme overkill. Hell there are things that a 600MHz G4 is over kill for.
    The issue is more along the lines of which chip has the greatest potential and will more than likely rule the roost in a particular discipline in industry. These really are the only people presently that can appreciate the perf. majority of users running word won't see a diff between a G5 and G4 and G3. However industry where time is money, it is important. Regardless of todays perfoormance of the G5; whether it is overall *twice* as fast as the x86 equiv as apple would have you believe, or whether its actually neck and neck in a lot of stuff. When it does go ahead of the x86 in a benchmark it tends to do it in leaps and bounds in order of magnitude. For example some of the photoshop results, if you are a pro, then the G5 would seam to be the chip for you. If you are into digi vid editing, G5 can now do 7 streams with only a plugin. nearest xeon equiv can only manage about 4 or 5 on a pc equiv application.
    The point is really that, Opteron and Xeon (particularly Xeon) are running extremely optimized code these days what with SSE2 etc... support widespread in apps. The opteron very cleverly on amd's part does have room for extra performance, in 64 bit mode, but then so does the G5. In its 32 bit mode software is as optimized as it can get for it presently. The opteron is heavily based on the athlon, it uses the same fpu and int units as the athlon xp, it adds (on die memory controller) SSE2 extensions and x86-64 extensions. Its core units int, fpu and scheduler is essentially that of an athlon xp. SSE2 is widespread in software now thanks to intel needing to push its adoption so that the P4 could perform. Looking at SUSE AMD 64 which we run in work, the opteron gains about 10 ->15 percent for 64bit mode. Rumor has it that Pathway software are working with AMD on an extremely optimized 64 bit compiler for x86 64 mode. Supposedly it produces 40% improvements in a few tests with the majority of tests gaining 15% extra speed over a already heavily optimized GCC for x86.
    You see what i am getting at, is that presently the PPC970 is RADICALLY different to any PPC that came before. Architecturally the chip was built for extremely high performance and extreme parallelism. Case in point is the fact that it can issue so many instructions for execution, that it can retire so many intructions, that it has so many more registers, that it has dual COMPLEX fpu, integer units, that it can have 215 instructions in flight. These weren't just put there by IBM as a marketing tool. This chip was designed for real world performance. Having been to the IBM developer conf last year and talking to the designers present, i can assure you, they explicitly were targeting future performance. Its raw brute horse power is allowing the PPC970 to compete extremely favorably with x86 presently. But when you do use a compiler that can generate code that uses its extremely advanced scheduler to saturate the pipes, that can generate code that runs on dual int and dual fpu pipes in parallel you get literally double or tripple the performance of what you get from an equivalent x86 running with an optimized compiler. Currently as i already quoted ars on... PPC970 is best case scenario in the majority of time running at 50% efficiency. The IBM compilers are able to use all the bells and whisles which are present in the chip which are completely unexploited by GCC or any other compiler available for the G5 presently. Then there is the fact that the G5 incorporates a much much more advanced vector processing unit than the opteron or Xeon. It has the potential to improve performance when used in orders of significant magnitude. Current compilers including GCC have not and do not autovectorise like they do for x86.
    Furthermore historically x86 has had far better optimizations within GCC than PPC. the 970 being so new has had so little done for it within GCC that we are sitting on raw brute performance which is keeping us up with the fastest x86 money will buy.
    Now consider this. Chip technology this time around is different for Apple. Reason for this is their partnership with IBM. IBM are using the 970 in their own highend workstations and blade servers. IBM obviously are heavily supporting the G5 especially in light that they are offering a beta version of their compilers for Mac OS X. Apple never had that type of support from Motorola. Now consider that IBM has plans to incorporate autovectorisation in their compilers since their own blade servers will incorporate the altivec technology. Autovectorisation alone will yield massive performance increases.
    But lets forget the autovectorisation for a moment. Scientists at NASA recompiled their JET3D application with the IBM compiler and were shocked at the results.
    Integer scores rose nearly 80% over what GCC was giving them. Scaler fpu rose to over 210% what best case scenario GCC was giving them, and vector alitvec results rose 70% over what GCC was giving them.
    Other tests in the arstechnica forums reported similar experiences except in one or two cases... out of many extremely successful cases where XLF was 2 or 3 times faster than GCC, there were 2 cases where XLF and XLC was slightly slower than GCC. After a lot of debate in the forums these 2 anomolies were put down to the extremely beta status of the compiler.
    AFAIK we are still talking 32 bit mode PPC970 ;)
    Its widely known that theoretically the 970 has significantly higher performance than an equivalently clocked Opteron.
    When a proper compiler like the XLF and XLC are used this theoretical becomes reality.
    When GCC is used and "incompatible" or non efficient compiler is used it only operates at 50% peak theoretical performance.
    Todays "optimized apps" have a huge way to go in terms of tapping into the 970 in 32 bit mode let alone 64 bit mode ;) For a start start using the IBM compiler until GCC becomes better optimized. Thats not to dish on GCC, im sure it will improve and its extremely like that from here on in the performance we get out of 970 in each GCC revision will be better in leaps and bounds.
    The point is that its amazing that in a worst case scenario these days (GCC being used on legacy code) that the G5 is able to keep up with and beat in some instances the best that the x86 can throw at it.
    Now that OS X is unix based you can download the source yourself and compile with XLC or XLF in some cases (most modern apps are written in C so more than likely 90% of time you will be using XLC). If you do this you will see what im talking about. Performance in most cases doubles or tripples what GCC can do. Proves exactly what ars was saying. In my opinion raw brute horse power on legacy apps makes it extremely competetive with the fastest optimized x86 apps. Use the optimized equiv compiler for the 970 and you got the fastest PC full stop. Think if performance improves significantly with GCC by a recomile, imagine what you are getting when you recompile with XLC.
    On a side note, apple will probably have to redo their spec results. IBM will be publishing spec results for their blades when finalised and are likely to blow apples results out of the water, even though both use the same chip on similar hardware! It will look strange when an IBM dual 1.6 blade blows the nickers of a dual 2GHz G5 in spec fpu and spec int, all because apple used GCC! By the way incase you are wondering IBM compilers presently generate GCC compatible results and are eligible to be used for compiling SPEC.
    Now show me an x86 chip that has this potential performance waiting to be tapped under the hood. ;)

    Phew ... im wrecked ....end rant!
     
  14. superbovine thread starter macrumors 68030

    superbovine

    Joined:
    Nov 7, 2003
    #14
    well i_wolf if you read the max pc article they covered all those points.

    [mod. edit - Insult.]
     
  15. superbovine thread starter macrumors 68030

    superbovine

    Joined:
    Nov 7, 2003
    #15
    again if you read the article in max pc you could have just said, yes i agree with the article in max pc. they said in a lot less boring way too.

     
  16. Veldek macrumors 68000

    Veldek

    Joined:
    Mar 29, 2003
    Location:
    Germany
    #16
    Well, I found it very interesting!
     
  17. MacBandit macrumors 604

    MacBandit

    Joined:
    Aug 9, 2002
    Location:
    Springfield, OR (Home of the Simpsons)
    #17
    I just wish these magazines would get it through there head and do real tests. Come on how many people here run one task at a time? Why don't they run multiple app benchmarks at a time. That would surely show off not only a dual processor machine but OSX itself.
     
  18. agdickinson macrumors member

    Joined:
    Jun 17, 2003
    Location:
    Marston Green, West Midlands, UK
    #18
    I found the reply by i_wolf to be very informative, albeit a bit long :) .. Then again the reply had to be long (not a two sentance answer to do it right :O) ) :D

    I've got a G5 Dual 2Ghz and have found the performance to be amazing. It's very interesting.

    I've found in the PC world that many PC types think if it's not x86 it's not good, slow...

    The G5 is still a relatively new processor and like i_wolf say's many developers have not yet taken on board the new compilers which cater for the changes in the internal architecture of the G5 versus the earlier G3/G4 processors.

    I guess many programmers now rely on fast processors so that they don't have to do much tweaking of code. I think like i_wolf says things will get even better considering much of the code out there is no where near tuned for the G5.

    Though it still depends upon developers making use of newer facilities in compilers / coding technicques
     
  19. LethalWolfe macrumors G3

    LethalWolfe

    Joined:
    Jan 11, 2002
    Location:
    Los Angeles
    #19
    No one has posted any links or detailed info, but the Maximum PC article sounds a lot like the benches done by PCWord a few months agoLink. Are these the same benches?


    Lethal
     
  20. MacBandit macrumors 604

    MacBandit

    Joined:
    Aug 9, 2002
    Location:
    Springfield, OR (Home of the Simpsons)
    #20
    If a backlash is coming it will be in the January issue. What you read in a Magazine is always about 45-60 days behind what is actually happening due to print deadlines. For that reason all reader write-ins are almost always 2 months after the article was published.
     

Share This Page