i don't think we'll be hearing from this guy for a very long time.
he be banned.
so don't expect any prompt answers about this mythical consumer 64-bit version of windows due in the next three months.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
PowerMac G5: World's Fastest Desktop Computer
- Thread starter MacRumors
- Start date
- Sort by reaction score
- Status
i don't think we'll be hearing from this guy for a very long time.
he be banned.
so don't expect any prompt answers about this mythical consumer 64-bit version of windows due in the next three months.
ECC is for those types of applications where you really really do not want any alpha particles changing a bit. That would be for high-flying computers such as the ones found on combat aircraft, and maybe cruise missiles. In this, it really is a matter of life or death.
ECC is also for databases such as the ones used by banks. Imagine a wrong bit is set or cleared, and you can imagine the shock of people when they learn that their $50K savings account is $5 instead.
ECC only adds one bit per byte, and is akin to the parity bit of yesteryear. ECC uses these extra bits to correct single bit errors, or detect multi-bit errors. That is all. There is nothing magical about ECC and large memory sizes.
As far as using registered memory, thats only there to give the hardware extra time to talk to the relatively slow DRAM chips. Registered DDR DRAM is more expensive than unbuffered DDR DRAM. You have to pay for the registers. Its actually good that Apple uses unbuffered DDR DRAM. It makes it cheaper to upgrade memory since its more common.
[humor]
Point? I thought that you were just using the thread to single-topically hit the 3,000 post mark?[/humor]
-Richard
Yeah, I was going to join in, but it was much more fun watching the CAPS ratio starting to climb, knowing that Rower_CPU was lurking.
If that's wrong, I don't want to be right.
Also, I er...well, to be honest, I didn't have a clue what you were talking about.
Basically I'm a bit thick.
If it ain't cricket or mars bars, I'm pretty much out of the loop.
Can someone at least start a rant about *something* because I really want this thread to pass 500 posts?
Go team!
Brother Mugga
i've played rugby and soccer (football
) when i lived in the UK, but cricket was the one sport i never, ever understood.but thats another thread for another time.
Well, when you're in, you try not to get out and...
And certainly one that would take us over the 500 posts NO SWEAT.
Incidentally, England covered themselves in glory today by losing to ZIMBABWE.
Oh yeah, we're *pre-tty* cool, I can tell ya...
Brother Mugga
PS: While at college I arranged a Germans v The Rest Of The World cricket match (Subtext? What's a subtext?). About four Yanks played (really well, in fact - two of them were semi-pro baseball players; MAN could they throw). Just when it looked like the Germans were about to lose, they bowled out the last four batsmen for 2 runs and clinched what I believe is referred to as 'a thrilling victory'.
Anyone who thinks Germans aren't emotional really needed to check out the four-lap celebrations that ensued.
Sorry to clog up the thread with this, but it really was one of the funiest things I've ever seen. They were *so* sweet (you know, in that teutonic way of theirs, bless 'em).
Originally posted by job
i've played rugby and soccer (football
but thats another thread for another time.
And certainly one that would take us over the 500 posts NO SWEAT.
Incidentally, England covered themselves in glory today by losing to ZIMBABWE.
Oh yeah, we're *pre-tty* cool, I can tell ya...
Brother Mugga
PS: While at college I arranged a Germans v The Rest Of The World cricket match (Subtext? What's a subtext?). About four Yanks played (really well, in fact - two of them were semi-pro baseball players; MAN could they throw). Just when it looked like the Germans were about to lose, they bowled out the last four batsmen for 2 runs and clinched what I believe is referred to as 'a thrilling victory'.
Anyone who thinks Germans aren't emotional really needed to check out the four-lap celebrations that ensued.
Sorry to clog up the thread with this, but it really was one of the funiest things I've ever seen. They were *so* sweet (you know, in that teutonic way of theirs, bless 'em).
Ummm ... 10.2.7 support 64-bit memory addressing.
That would be a consumer OS for a 64-bit computer.
Panther is *optmized* for 64-bit apps. 10.2.7 supports 64-bit apps.
In view of all the disagreement over the benchmarks, this little piece over at MacFixit seems interesting
I don't know what these guys reliability record is though, so i'm not sure that they're not talking out of their asses
Would be quite interesting if this is true.
I don't know what these guys reliability record is though, so i'm not sure that they're not talking out of their asses
Would be quite interesting if this is true.
1) 32 "bitness" can physically address 4 GB of memory and up to 36 GB of memory with PAE although it's still restricted to 4 GB/thread and their are some penalties.
2) 64 "bitness" provides alot more than just greater physical addressing, it's just that these advantages can't be realized in your standard fare of desktop programs. I recommend you read the Opteron review done by Ace's Hardware provided by the link below:
http://www.aceshardware.com/read.jsp?id=55000251
I have to respectfuly disagree with your contention that the PowerMac is pro-sumer. To cover the consumer field Apple has the eMac & IMac lines. The Power Mac was focused on the professional market. I'm not sure what the G5 is focused on, but it certainly doesn't meet the needs of the average professional. Expandability is the key to the professional market, the G5 has very little to offer besides PCI slots.
I have a hard time believing that anybody believes that the consumer is going to shell out for an underachiving machine.
Dave
I have a hard time believing that anybody believes that the consumer is going to shell out for an underachiving machine.
Dave
Why can you compare the G5 to Intel processors but not AMD processors? Could it be because GCC results for the Opteron show that it is actually stronger SPEC-wise than the G5?
And I agree, GCC is a UNIX compiler, OSX is also an UNIX based OS, GCC is also not "streamed" meaning that the Pentium 4 scores will be significantly poorer than with say ICC or BSD compiler.
Not that it matters that much considering that even with the GCC compiler, the Pentium 4's integer performance is stronger than the G5's (11% higher SPECint if I remember my previous post right).
Most of the FP tests were written using FORTRAN, which would naturally require a Fortran compiler. In this particular comparison, a Nagware compiler was used which I find strange considering GNU's (yes the same makers of the GCC compiler) Fortran 77 compiler is more widely used and well known (some of the reasons used to justify the use of the GCC compiler) as well as being a better compiler altogether. Perhaps what we should be focusing on is the Nagware compiler used.
Hi Rower;
I have to disagree with your with repsect ot the expandability issue. First. many professionals will tie up the firewire ports doing other things. In any event external storage has its place but is not a replacement for internal storage. Its not the size of the storage but the flexibility in configuration that is the issue.
The basis for the performance sucks statement is not immflammatory, it is based on Apple posted specs. The integer throughput is terrible on these machines. Granted that is not everything when it come to machine behaviour. The only thing that showed a signnificant performance increase, an a MHz adjusted basis, was floating point. This is great if your doing heavy floating point, but probally won't help many other tasks.
As far as ram goes you may have a point if the preceived market is the consumer. I would not expect such addressing in an iMac for example. However for some of Apples target markets it certainly is a concern.
As far as pricing goes I've said before Apple's software is worth a little extra, but not thousands. The G5 especially at the lower end, where all the features of the high end disappear, are just over priced.
I do not believe that my conclusions are ill founded at all. They are obviously the results of differrent metrics than others are using. The utility of these machines will be defined more by Panther than their CPU performance.
Your also mistaken if you think that Apple can do nothing to make me happy. That is not the case at all. I'm just surprised at the glowing reception this machine has received, it is not warranted.
Thanks
Dave
I have to disagree with your with repsect ot the expandability issue. First. many professionals will tie up the firewire ports doing other things. In any event external storage has its place but is not a replacement for internal storage. Its not the size of the storage but the flexibility in configuration that is the issue.
The basis for the performance sucks statement is not immflammatory, it is based on Apple posted specs. The integer throughput is terrible on these machines. Granted that is not everything when it come to machine behaviour. The only thing that showed a signnificant performance increase, an a MHz adjusted basis, was floating point. This is great if your doing heavy floating point, but probally won't help many other tasks.
As far as ram goes you may have a point if the preceived market is the consumer. I would not expect such addressing in an iMac for example. However for some of Apples target markets it certainly is a concern.
As far as pricing goes I've said before Apple's software is worth a little extra, but not thousands. The G5 especially at the lower end, where all the features of the high end disappear, are just over priced.
I do not believe that my conclusions are ill founded at all. They are obviously the results of differrent metrics than others are using. The utility of these machines will be defined more by Panther than their CPU performance.
Your also mistaken if you think that Apple can do nothing to make me happy. That is not the case at all. I'm just surprised at the glowing reception this machine has received, it is not warranted.
Thanks
Dave
Re: Hmmmm and again I say Hmmmmm
I have a suspicion that with time and the clearing of the air, more people will realize that the points are valid. So soon instead of being behind Rover you will step into the glow of enlightenment and true question some of the design issues with this machine.
About the case, its rather nice but needs a few more holes in it. Like one or two for externally accessible media, and a couple of internal holes for disk drives. I know that some believe the wave of the future is to clutter your desktop with a rats nest of wires running to differrent external devices, but it is not my future! So yes it is a nice but wanting housing for the G5.
Armari is not familiar to me. But if it is like many Intel work stations it probally comes with a great deal of expansion capability. Youmight end up missing that.
The clincher is Panther as far as operating systems go. That may have more of a impact than these machines.
One good thing that I have not seen considered is that these machines should ramp quickly. If Apple can put a 2.5GHz processor in this machine well before Christmas they may have accomplished something. That would fit in nicely with the statement that we will see 3GHz in 12 months. Then a gains 3GHz for Christmas would be nice.
Dave
I have a suspicion that with time and the clearing of the air, more people will realize that the points are valid. So soon instead of being behind Rover you will step into the glow of enlightenment and true question some of the design issues with this machine.
About the case, its rather nice but needs a few more holes in it. Like one or two for externally accessible media, and a couple of internal holes for disk drives. I know that some believe the wave of the future is to clutter your desktop with a rats nest of wires running to differrent external devices, but it is not my future! So yes it is a nice but wanting housing for the G5.
Armari is not familiar to me. But if it is like many Intel work stations it probally comes with a great deal of expansion capability. Youmight end up missing that.
The clincher is Panther as far as operating systems go. That may have more of a impact than these machines.
One good thing that I have not seen considered is that these machines should ramp quickly. If Apple can put a 2.5GHz processor in this machine well before Christmas they may have accomplished something. That would fit in nicely with the statement that we will see 3GHz in 12 months. Then a gains 3GHz for Christmas would be nice.
Dave
Agreed - needs more drive space
I agree with many here, the thing is within an inch of the Dell they were pitting it against on stage and over 3 inches talled than the MDD G4 tower yet it has 2 less 5.25 drive bays than the Dell?!?!
I don't think it would have killed them to make it a couple inches taller to fit a 2nd optical bay and space for a pair more of HD's.
(But what do I know - maybe there's some industry standard that says 21" is the bigest tower you should ever make. )
That said, I love this machine and personally don't need more PCI slots or HD's that I couldn't use FW800 or a SCSI card for. I'd buy a Dual 2Ghz right away if I had the cash!
I agree with many here, the thing is within an inch of the Dell they were pitting it against on stage and over 3 inches talled than the MDD G4 tower yet it has 2 less 5.25 drive bays than the Dell?!?!
I don't think it would have killed them to make it a couple inches taller to fit a 2nd optical bay and space for a pair more of HD's.
(But what do I know - maybe there's some industry standard that says 21" is the bigest tower you should ever make.
)That said, I love this machine and personally don't need more PCI slots or HD's that I couldn't use FW800 or a SCSI card for. I'd buy a Dual 2Ghz right away if I had the cash!
For the last time, wizard, do a side-by-side price comparison of the G5s and a 64-bit PC with equivalent features. I've asked twice and you seem to have ignored me both times.
If you can prove your point, I will shut up; if you can't/won't I suggest you do likewise.
If you can prove your point, I will shut up; if you can't/won't I suggest you do likewise.
Comments suggesting that "integer throughput is terrible" on the G5 is just ridiculous.
A single 2.0 GHz G5 is only 10% slower in Apple's SPECint test than a 3.0 GHz P4. So you're saying that any SPECint scores below the 3.0 GHz P4 are "terrible"? According to that, any P4 at 2.7 GHz or below would then have the same "terrible" throughput as the 2.0 GHz G5. Do you really believe that?
The PC world touts all of the SPEC scores about the Athlon/Opteron/P4/Xeon that are higher than the G5, especially the P4/Xeon scores that they say "prove" Apple rigged the results. Apple has clearly explained how they did their test, and their logic still remains unrefuted. At best, people can only say they don't LIKE the way Apple did the test. But not liking their test doesn't INVALIDATE the results of that test.
Some have intelligently mentioned that IBM showed better SPEC scores for a G5 at 1.8 GHz than Apple's results on the 2.0 GHz G5. So to believe Apple "rigged" the test is to suggest that Apple INTENTIONALLY tried to make the 2.0 GHz G5 look 10% slower than a P4 in SPECint, and 20-30% slower than IBM's previously published 1.8 GHz results.
"Yeah...what a great idea! Let's design a test that makes a 2.0 GHz chip look slower than a 1.8 GHz chip, and STILL get beaten by a P4 in SPECint!" It's ludicrous to even suggest it. You don't rig a test to make your own scores look lower...NEVER. All Apple had to do was test the 2.0 GHz G5 the same way IBM tested the 1.8 GHz G5, and Apple's scores would have been 11% higher than IBM's previously published results. So how can you explain the discrepancy?
There is only ONE logical answer to this question, no matter how many ways the Wintelers try to distort these results. The facts remain the same...Apple designed a test whose GOAL was to run the SPEC benchmark WITHOUT giving EITHER SIDE a specific advantage through optimizations.
It's that simple. It doesn't matter if YOU don't like that test...that is the test. The results of the test, using that formula and design, are completely valid. They aren't wrong, or lies, just because you would have done the test differently. Why is this so hard for some people to understand?
If Apple did anything that was against the SPEC specifications, especially when it's already generated so much controversy, why hasn't the SPEC organization come out with a press release saying so? Because Apple didn't do anything underhanded. If anything, they penalized their OWN scores through their attempt to be fair.
Sure the Wintelers are making accusations, but those accusations are all based entirely on the wrong questions. So no matter HOW many times you keep asking those SAME questions, the results will not change. Apple's test remains firmly grounded in logic... It doesn't matter how you personally feel about it.
For their NEXT test, Apple should do a "ball's out kick ass" benchmark where each side can pull any optimizations they want to trick out the scores. I have no problem with that. And when they do, the scores will absolutely change...on both sides. But THAT was not the design goal of THIS test.
Maybe it's very clear to me because I've been working with computers and software development for over 15 years. But what seems so blatantly obvious to me is apparently not to the "wealth of experts" I've read over the past 3 days.
-- Ensoniq
A single 2.0 GHz G5 is only 10% slower in Apple's SPECint test than a 3.0 GHz P4. So you're saying that any SPECint scores below the 3.0 GHz P4 are "terrible"? According to that, any P4 at 2.7 GHz or below would then have the same "terrible" throughput as the 2.0 GHz G5. Do you really believe that?
The PC world touts all of the SPEC scores about the Athlon/Opteron/P4/Xeon that are higher than the G5, especially the P4/Xeon scores that they say "prove" Apple rigged the results. Apple has clearly explained how they did their test, and their logic still remains unrefuted. At best, people can only say they don't LIKE the way Apple did the test. But not liking their test doesn't INVALIDATE the results of that test.
Some have intelligently mentioned that IBM showed better SPEC scores for a G5 at 1.8 GHz than Apple's results on the 2.0 GHz G5. So to believe Apple "rigged" the test is to suggest that Apple INTENTIONALLY tried to make the 2.0 GHz G5 look 10% slower than a P4 in SPECint, and 20-30% slower than IBM's previously published 1.8 GHz results.
"Yeah...what a great idea! Let's design a test that makes a 2.0 GHz chip look slower than a 1.8 GHz chip, and STILL get beaten by a P4 in SPECint!" It's ludicrous to even suggest it. You don't rig a test to make your own scores look lower...NEVER. All Apple had to do was test the 2.0 GHz G5 the same way IBM tested the 1.8 GHz G5, and Apple's scores would have been 11% higher than IBM's previously published results. So how can you explain the discrepancy?
There is only ONE logical answer to this question, no matter how many ways the Wintelers try to distort these results. The facts remain the same...Apple designed a test whose GOAL was to run the SPEC benchmark WITHOUT giving EITHER SIDE a specific advantage through optimizations.
It's that simple. It doesn't matter if YOU don't like that test...that is the test. The results of the test, using that formula and design, are completely valid. They aren't wrong, or lies, just because you would have done the test differently. Why is this so hard for some people to understand?
If Apple did anything that was against the SPEC specifications, especially when it's already generated so much controversy, why hasn't the SPEC organization come out with a press release saying so? Because Apple didn't do anything underhanded. If anything, they penalized their OWN scores through their attempt to be fair.
Sure the Wintelers are making accusations, but those accusations are all based entirely on the wrong questions. So no matter HOW many times you keep asking those SAME questions, the results will not change. Apple's test remains firmly grounded in logic... It doesn't matter how you personally feel about it.
For their NEXT test, Apple should do a "ball's out kick ass" benchmark where each side can pull any optimizations they want to trick out the scores. I have no problem with that. And when they do, the scores will absolutely change...on both sides. But THAT was not the design goal of THIS test.
Maybe it's very clear to me because I've been working with computers and software development for over 15 years. But what seems so blatantly obvious to me is apparently not to the "wealth of experts" I've read over the past 3 days.
-- Ensoniq
Sorry if this has already been posted
http://www.xlr8yourmac.com
SPEC ....SMREK...that's a Azz Kickin folks. If a Dual G5 is twice as fast as a Dual 2.6 Xeon then their claims of it beating a Dual 3.06Ghz Xeon should be spot on.
Funny how everyone points to the Haxial site as if the guy actually has credentials LOL hahahahah muahahahah please. He was blown out the water by Greg Joswiaks report. Misleading.....bah... . Wintels domination in speed is over. Apple has already said we'll be a 3 Ghz in 12 months. That will keep up with anything Intel has planned including Prescott and Tejas.
http://www.xlr8yourmac.com
SPEC ....SMREK...that's a Azz Kickin folks. If a Dual G5 is twice as fast as a Dual 2.6 Xeon then their claims of it beating a Dual 3.06Ghz Xeon should be spot on.
Funny how everyone points to the Haxial site as if the guy actually has credentials LOL hahahahah muahahahah please. He was blown out the water by Greg Joswiaks report. Misleading.....bah... . Wintels domination in speed is over. Apple has already said we'll be a 3 Ghz in 12 months. That will keep up with anything Intel has planned including Prescott and Tejas.
No to wires.
r.e. Wizard's post
I certainly agree with you on the 'no to wires snaking all over the place' front. In which context I also agree that removing the option for the second optical drive may be a mistake. This is far more important (in my mind) than the internal HD issue; with serial ATA giving you 500Gigs, I really don't see what people are worried about. If you need more than that in a machine then you're WAY beyond consumer and perhaps even pro-sumer.
However, having got used to two OD, the lack of a second might be more of a bind. Someone on another website suggested it might be part of a deal on getting iTMS up and running? (record companies having long been down on twin ODs). However, given the ease with which anyone who want's to rip CDs can plug in an external OD, I think this unlikely.
I guess we'll just have to wait and see how things pan out once we get to use them. This is one of the reasons I'm not going to order one until Christmas - to see what the gripes might be (that and the hope that their lineup will then comprise single 1.8, dual 2.0 and dual 2.5 with comensurate price drops for the 2.0 Gig).
To be honest, though, it's just nice that a genuine opportunity to 'switch' to Apple has now opened up for people (like me) who are interested in workstation performance.
Brother Mugga
PS: I've mentioned this on another site, but...well, is it just me or is there a vaguely 'Art Deco' styling to the new case? (which is probably why I like it). Seriously, have another look...
r.e. Wizard's post
I certainly agree with you on the 'no to wires snaking all over the place' front. In which context I also agree that removing the option for the second optical drive may be a mistake. This is far more important (in my mind) than the internal HD issue; with serial ATA giving you 500Gigs, I really don't see what people are worried about. If you need more than that in a machine then you're WAY beyond consumer and perhaps even pro-sumer.
However, having got used to two OD, the lack of a second might be more of a bind. Someone on another website suggested it might be part of a deal on getting iTMS up and running? (record companies having long been down on twin ODs). However, given the ease with which anyone who want's to rip CDs can plug in an external OD, I think this unlikely.
I guess we'll just have to wait and see how things pan out once we get to use them. This is one of the reasons I'm not going to order one until Christmas - to see what the gripes might be (that and the hope that their lineup will then comprise single 1.8, dual 2.0 and dual 2.5 with comensurate price drops for the 2.0 Gig).
To be honest, though, it's just nice that a genuine opportunity to 'switch' to Apple has now opened up for people (like me) who are interested in workstation performance.
Brother Mugga
PS: I've mentioned this on another site, but...well, is it just me or is there a vaguely 'Art Deco' styling to the new case? (which is probably why I like it). Seriously, have another look...
SPEC int and fp tests CPU speed, not throughput, and I don't recall hearing any respectable member on this forum saying (or suggesting for that matter) that integer performance is terrible. Personally, I think it's very impressive looking at IBM's and Apple's scores.
No, it doesn't invalidate the test, the way that Apple conducted the test would invalidate a fair comparison and their logic is inherently flawed in assuming that two GCC compilers are equal when they are clearly different, the only common aspect they possess is in name only. Not to mention that the three systems compared weren't all configured on even grounds.
Assuming linear scaling and a slightly faster bus a 2.0 GHz G5 compiled with Visual Age would, under the best scenario score around an 1060 in SPECint and 1200 in SPECfp, it would require a score of nearly 1500 in SPECfp in order to match Apple's claims.
Apple's Pentium 4 was 75% slower than Intel's Pentium 4 in FP performance and 30% slower in integer performance. The SPEC rates are even worse. As far as I can tell, Apple's G5 isn't going to score anywhere near 75% slower than the 2 GHz PPC970 in FP.
Absolute BS, x86 GCC and PPC GCC are two entirely different compilers, the Nagware FORTRAN compiler is definitely not the best choice for a "fair" comparison, never mind that the Pentium 4's compiler flags were not the best choices for speed, and the G5 was using an high performance malloc library optimized for speed and perfectly suited for single threaded SPEC, a command called hwprefetch -8 that disables software prefetching and enables up to eight hardware prefetch streams, and modified CPU registers to enable memory read bypass, reducing the time required to an read request.
It's quite simple, they did their little comparison, and didn't submit the scores (and for good reason) to SPEC, SPEC only looks at submitted scores and only those that are done credibly (SPEC looks over all the test settings and configurations to make sure of they are done credibly and performs all sorts of tests to confirm these settings) are made official SPEC submissions (the ones on their website http://www.spec.org).
It seems that you don't realize that doing the test differently would change the results entirely and that the results have no value unless they are compared in an fair way requiring an fair comparison in the first place.
Personally, I prefer looking at official SPEC submissions, multiple third party reviews, technical white papers and if possible, some personal experience to determine how well an cpus performs instead of getting my information from the company thats trying to market and sell those cpus but hey, thats just me.
I agree 120%. Who actually believes that Apple is any different then any other mainstream corp. They will do whatever it takes to make themselves smell good and come out on top. I don't question this at all. Personally I don't buy the computers for the symbol on the front of them or the products because of who makes them. I buy them for the reasons of quality and user experience etc..
The simple fact is that the computer industry is a very very competive thing and you can't believe any manufacturors claims to be 100% correct. In my oppinion though the simple fact that these claims are being taken so seriously and causing so much controversy leads me to believe that Apple is on to something and these machines have really touched a sore spot. I mean it's caused all but an internet riot among the computer geeks. What would it take to do that?
I personally am waiting for the 3rd party head to head comparrisons especially the ones from the major PC magazines and internet sites as those will be the ones that will try to distort the advantage as much as possible. If Apples machines show an advantage at all in those tests then yes we have a really fast machine on our hands until then let's just sit on our hands and wait. Enough bickering.
That is exactly what I'm trying to say! You have to remember that the P4 is very old technology so to fail against that machine is a problem. Granted integer performance is not high on everybodies list of important features, but it is pretty terrible that a brand new processor design can not beat an aged chip. On a MHz ajusted basis the G5 apparently does not even beat the G4, pretty terrible. But one shold understand that that doesn't mean that other performance parameters for the chip are not competitive, the float performance is outstanding.
I think Apple did a good job with the tests and have been honest and up front about the tests. In the end, after all of the excitement wanes, I believe the community will ocme to that conclusion.
My whole point, maybe I haven't gotten it across well, is that the machine should really be compared to technology from Intel and AMD just coming on the market. It is here that the G5 will be shown to be lacking. Comparing the G5 to yewstedays PC hardware may have been required, after all you can't expect the competition to give up their latest and greatest before it hits the market for bench marking, but that is not a way to judge the machine in the market. You have to compare it to machines of the same generation. By that measure the G5 barely maintains parity.
Thanks
Dave
I think Apple did a good job with the tests and have been honest and up front about the tests. In the end, after all of the excitement wanes, I believe the community will ocme to that conclusion.
My whole point, maybe I haven't gotten it across well, is that the machine should really be compared to technology from Intel and AMD just coming on the market. It is here that the G5 will be shown to be lacking. Comparing the G5 to yewstedays PC hardware may have been required, after all you can't expect the competition to give up their latest and greatest before it hits the market for bench marking, but that is not a way to judge the machine in the market. You have to compare it to machines of the same generation. By that measure the G5 barely maintains parity.
Thanks
Dave
Lies, Damn Lies, and Benchmarks
This is bull****. GCC is an open-source compiler which makes it open to scrutiny. If anyone puts in benchmark-specific optimizations, it would be immediately evident and removed. (Am I only person around who still remembers the LibMoto library that would make your G3/G4 sing on the FPU benchmark but would crash your software because the arctan lookup tables were not sufficiently precise?) Can ICC claim this scrutiny? Or, for that matter, can CodeWarrior?
Second, GCC is geared toward portability. It is constructed into two parts. The head end (where most of Apple's gcc commits are) benefit all compilers, the back-end gets tailored to specific compilers. Since the back end for the x86 has over a decade of development, while the PPC backend was in a sorry state until recently, if anything it would skew the benchmark away from Apple.
Third, GCC is used in a lot of scientific/workstation work. This is because often (but not always) the extra development time of optimizing for a particular compiler (or the cost of purchasing an optimized compiler) is not justified. For general application performance, it would be "fairer" to compare Microsofts VC compiler with CodeWarrior, but Apple's benchmark does still have relevance to scientists and researchers--what relevance does standard practice of reporting peak SPEC have to anyone but people comparing p***s sizes? BTW, I believe that MVCC would have looked much better relative to CW because one is a mature compiler and the other might not even work with the G5!
Look, it is correct to say "using GCC doesn't 'normalize out' the compiler as Apple claims" and it is correct to say "standarizing on a single compiler is not standard practice when reporting SPEC scores" and its correct to say "`the worlds fastest personal computer' is a bunch of marketting hoopla designed to evoke just the sort of repsonse in the press it has gotten", but it is not correct to claim "the only common aspect GCC possess across platform is in name only".
Of course, I don't know why I bother responding. If your purchasing decisions are based on SPEC scores and not real application performance, I suggest you go troll among the other ignoramuses at AMDZone and Overclockers. And please, take some of those "Rah Rah! Apple is the greatest" Mac zealots along with you.
The "rest of the us" will be getting real work done on our Pentium 4 Xeons, G5/970's and Opterons.
[Note: On re-reading your pos, I noticed you aren't necessarily basing your purchasing decision on SPEC. Sorry for the insinuation.]
This is bull****. GCC is an open-source compiler which makes it open to scrutiny. If anyone puts in benchmark-specific optimizations, it would be immediately evident and removed. (Am I only person around who still remembers the LibMoto library that would make your G3/G4 sing on the FPU benchmark but would crash your software because the arctan lookup tables were not sufficiently precise?) Can ICC claim this scrutiny? Or, for that matter, can CodeWarrior?
Second, GCC is geared toward portability. It is constructed into two parts. The head end (where most of Apple's gcc commits are) benefit all compilers, the back-end gets tailored to specific compilers. Since the back end for the x86 has over a decade of development, while the PPC backend was in a sorry state until recently, if anything it would skew the benchmark away from Apple.
Third, GCC is used in a lot of scientific/workstation work. This is because often (but not always) the extra development time of optimizing for a particular compiler (or the cost of purchasing an optimized compiler) is not justified. For general application performance, it would be "fairer" to compare Microsofts VC compiler with CodeWarrior, but Apple's benchmark does still have relevance to scientists and researchers--what relevance does standard practice of reporting peak SPEC have to anyone but people comparing p***s sizes? BTW, I believe that MVCC would have looked much better relative to CW because one is a mature compiler and the other might not even work with the G5!
Look, it is correct to say "using GCC doesn't 'normalize out' the compiler as Apple claims" and it is correct to say "standarizing on a single compiler is not standard practice when reporting SPEC scores" and its correct to say "`the worlds fastest personal computer' is a bunch of marketting hoopla designed to evoke just the sort of repsonse in the press it has gotten", but it is not correct to claim "the only common aspect GCC possess across platform is in name only".
Of course, I don't know why I bother responding. If your purchasing decisions are based on SPEC scores and not real application performance, I suggest you go troll among the other ignoramuses at AMDZone and Overclockers. And please, take some of those "Rah Rah! Apple is the greatest" Mac zealots along with you.
The "rest of the us" will be getting real work done on our Pentium 4 Xeons, G5/970's and Opterons.
[Note: On re-reading your pos, I noticed you aren't necessarily basing your purchasing decision on SPEC. Sorry for the insinuation.]
Hi Rover;
I have no intention of going shopping for somebody else. It is really up to the purchaser to determine the make up of his or hers machine. Assuming one knows how to use the web, Opteron based machines can be found using AMD or NVIDIA chipsets. I will continue to ignore your fishing excercise, you should know as well as I what is on the market, if you don't you really have no business responding to this thread.
What is clear is that people should not be going out and buying a G5 based on assumed performance. Specific applications yes, OS/X yes, a cute enclosure well maybe, but not because they believe that it is an outstanding machine performance wise. Performance will have to be judged against its (G5) contemporaries. Lets face it Panther is a much more compelling reason to buy a G5 than its much touted performance.
Thanks
Dave
I have no intention of going shopping for somebody else. It is really up to the purchaser to determine the make up of his or hers machine. Assuming one knows how to use the web, Opteron based machines can be found using AMD or NVIDIA chipsets. I will continue to ignore your fishing excercise, you should know as well as I what is on the market, if you don't you really have no business responding to this thread.
What is clear is that people should not be going out and buying a G5 based on assumed performance. Specific applications yes, OS/X yes, a cute enclosure well maybe, but not because they believe that it is an outstanding machine performance wise. Performance will have to be judged against its (G5) contemporaries. Lets face it Panther is a much more compelling reason to buy a G5 than its much touted performance.
Thanks
Dave
What's so "old" about the P4? The P4 they benchmarked against is a great (new) design using the same process (130nm) as the G5. Just because it has a lot of silicon wasted on legacy doesn't make a chip "old technology".
If you knew anything about the way the P4 is structured, you'd be surprised, as I was, that the single CPU G5 achieved anything close to SPEC integer parity with the P4 (actually, not that surprised because you would have been aware of IBM's benchmarks in Microprocessor Forum last year).
Finally, I advise you not to talk in a "megahertz adjusted basis" ever again, because your PC-loving friends will laugh you out of the room. Such commentary is Apple-marketting smoke-and-mirrors designed to distract us from the fact the G4 had lost parity with the x86 (Athlon and P4). Now that Apple finally has a real workstation-class processor, we can finally shed ourselves of that embarrassment.
(The only time such a thing is close to relevant is in notebooks. In those cases, it is better to look at the amount of work done / watt instead.)
Yes, it is true that the G4 (and G5) do more work per clock cycle than the P4. But the P4 is an engine that is designed to rev higher--that's like saying a PT Cruiser is faster when at the same RPMs as a Honda S2000, so therefore it makes it a faster car...
Apple hasn't released any G4 benchmark done under the same conditions. Most likely because it performs so horribly in SPEC as to be attrocious. Also, you cannot get a G4 at 2 GHz yet so the whole argument is moot. When you can, then we can compare it to the G5s going at 3 Ghz because by then, it surely must.
Take care,
terry
Common you PC guys. This thread has disintegrated into BS on top of BS. Bogus price comparisons and everyone's a professor in CPU architecture, with a book due out soon.
If you look at the G5's competition, and consider the pro graphics/video market it is designed for, it has no competition in the marketplace on a performance for value equation.
The DP 2Ghz G5 is cheaper than that Opteron Boxx thing, I looked at the site myself and other have made detailed posts to this effect. And for floating point stuff like time consuming Pshop, video rendering etc, it pees all over Zeon and Pentium. This is when performance really counts. It's also much cheaper than dual Zeon or Pentium as well as the Boxx.
On top of that you have an unequalled OS that is about to get better, and make the system faster and we can expect considerable Speed bumps within 6 months - so that should shut people up if a faster competition comes out sooner - you always have to consider that R&D of various companies is always out of sink.
With all due respect, I think you PC guys have got sucked in by Apple's marketing. This kind of interest/furore is exactly what they probably wanted, so people would start to take notice of the new PMac. Let's face it, if Apple did just release the world's fastest computer, who in the PC world really cares anyway. Speed doesn't make most people switch one way or the other. It's more a question of what do you want to do and what computer does it best, AND personal preferences - don't forget they ARE valid too. Many people in this thread seem to have forgotten that. So they put out a new slogan and all of a sudden PC guys are talkin about it. Do Apple really care if it's true? Everything moves so fast it could only be true for 2 or 3 months anyway. Brilliant marketing - and most of it's free.
I think instead of worrying about Apple's marketing, we should be focussing on the fact that, in spite of all the arguments, Apple finally again has a competitive price to performance equation (for the top machine anyway - not sure about the 1.6) and this frees customers to focus on the important issues which have been derailed over the last few years of megahertz wars (Brilliant Intel marketing strategy by the way). Ie What do you want to do and what machine does it best, and what are your personal preferences.
Personally, I think PCs are S**t.
Sorry couldn't resist a joke after all that.
- Status