It's the real world tests. Get mac app, get PC, face off.
It's pretty clear that the top end PPC now handily smokes the top end wintel, untill we see some Athlon64 numbers... Still, if I'm spending 2000-3000 on a photoshop or film or 3d animation rig, then the PM G5 gets my money. In the 1000-1500 range it's a harder call, and the PC offers more bang for the buck, but at 1299, the single 1.25Ghz G4 with L3 is not too shabby either.
Um....the PC's were SLAUGHTERED in the bake-offs (who'd a figured?). People make living running the programs used in THOSE tests. How many people make a living flogging a machine through a suite of benchmak tests?
Sounds like sour grapes. Apple (finally) has a no-nonsense chip with room to grow and some people's fragile worlds are collapsing.
SPECint rate and SPECfp rate scores that Apple provided for the G5 were 2P scores._ The P4 of course could not be tested in 2P mode, but the Xeon was._ Things look good for Apple's own SPEC benchmarks of the Xeon and P4 until you pan up at the official SPEC scores._ So which one is correct?_ The Apple scores or the ones on SPEC's site._ My money is with SPEC._ The difference is ridiculously large, and far too much for me to believe without some form of evidence._ There was no 1.7GHz Power4 to compare against in this test, but the 2P Opteron blows away the faster MHz G5.
So what is the true story of G5 performance?_ It certainly greatly improves on the G4, but I doubt we will ever find that out from Apple what true performance against modern PCs is. When they trotted out Adobe as a benchmark I tuned out._ Their ability to bend Photoshop benchmarks to suite their performance needs is legendary._ I wanted to hold out hope for Apple's PowerPC offering, but Jobs lies and stretches do a disservice to IBM and their powerful CPU._ This type of show from Apple may keep them pegged at 1.9% marketshare, but the sweet spot for real desktops is firmly below the$1,000 mark._ The PowerPC G5 system will allow Apple to hang onto what is left of their share, but they throw logic out the window with their over the top presentation that misleads, distorts facts, and leaves serious questions about actual performance._ Apple would do well to bring back clones and give Jobs the boot._ Here's hoping they do so they can regain marketshare and offer real competition to Microsoft, and further pressure on Intel._ Hand the real thanks to IBM and all of the borrowed PC technology._ That is what is allowing Apple to survive.
Those bench marks looks convincing to me, Apple Cheated on their specs and bechmark tests and I feel deceived by a cheat slight of hand dog and pony show carnival trick! (in the keynote demonstration)
It's the real world tests. Get mac app, get PC, face off.
It's pretty clear that the top end PPC now handily smokes the top end wintel, untill we see some Athlon64 numbers... Still, if I'm spending 2000-3000 on a photoshop or film or 3d animation rig, then the PM G5 gets my money. In the 1000-1500 range it's a harder call, and the PC offers more bang for the buck, but at 1299, the single 1.25Ghz G4 with L3 is not too shabby either.
[a little of topic again]
when i saw the prices of the G5 yesterday, i thought matsu would be pissed of and complaining about the $1999.00 pricetag of the low end. the 1.25Ghz G4 SP at $1299.00 would probally be $300.00 to much...
but that didn't happen.
so i will:
the lowend G5 is waaaaaaaaaaaaaaaaaaaaay to expensive, even the 1.8Ghz is a better option, no wait, the 2.0Ghz dp is the best deal...
oh well, wait the minute. back in august 2002, almost everybody who bought a powermac, bought the low end 867Mhz dp because that was the best bang for buck. now apple squeezed the specs of the low end G5 that much and give it a "confusing" pricetag of $1999.00. very close to the "better" option. in the netherlands the good/better difference is just 190 euroos. what would you buy? the better option i guess.
also this "high" pricetag of the low end gives lots of space for other products in the consumer matrix to find a comfortable place even though they're not (yet) upgraded to a G5
I do not know enough to comment on all the technical merits of this argument; I do however know, that reading the comments at the bottom of his critique made me, for the first time ever, thoroughly ashamed to be a Mac user.
Imagine, people who never posted before, suddenly posting about how apple lied... yet, they never really even paid attention to what kind of tests run and how they were compiled...... all you morons go FUD somewhere else.
I do not know enough to comment on all the technical merits of this argument; I do however know, that reading the comments at the bottom of his critique made me, for the first time ever, thoroughly ashamed to be a Mac user.
No need. The numbers at the SPEC site are skewed, 'unfair.' It's the way things go.
It is disappointing that Apple *appears* to have cheated somewhat on the spec test. I would expect better from Apple than say, Gateway. Who compared Quake 3 on a Gateway to Quake 3 running in Classic on an iMac.
Barto
I dont understand why you feel that Apple cheated on the SPEC test. Not that I give a wilted patunia's rear about SPEC test or any other benchmark out there. I dont run benchmarks on my computer, but rather I do work on it. Apple took a $4K Dell system and kicked its rear with a $3K system. It doesnt get any simpler than that as far as I am concerned.
Yawn. Spec junkies. Spec doesn't take SIMD units into account.
As Matsu says. Let's get the Apps and run'em. That'll tell you all you need to know. There are too many shady things going on for me to trust Spec %100
Well personally I don't believe any of the marketing crap spewed out by Apple OR Intel/AMD. I look more to independent websites/mags for my info. The problem on the mac side is finding an unbiased mac source (admitting that I'm relatively new to the mac side). All the mac magazines, aka macworld and macaddict, seem to just repeat what apple claims/says. Hell, reading macworld makes me think the writers work for apple (and why is it I get a subscription for just registering an apple product...i have like 4 subscriptions now) Anyway, the PC world is obviously more biased to the PC side but because the don't have to "fight" for the platform they can come out against stuff like the AMD rating or post specific benchmarks with real world apps providing the truth behind the AMD/intel debate.
These are all just observations. In the end, I don't care about who is the fastest because PC's don't run mac OS X, and that was the real star yesterday.
Intel uses their own especially for spec-tests tweakd compilers. Apple used the same compiler for every system. Apple is doing fair benchmarks.. the official spec-scores are quite misleading since every vendor supplies their own benchmarks, using whatever compiler they chose. AND.. they do not account for AltiVec.
Pixar uses Intel's especially for spec-tests tweakd compiler for Rendeman on x86, go figure... Face it, gcc is a slow compiler on both platforms, but Apples bigest crime in its number is useing tha NAGware compiler for the SPECfpu. Linux is alot slower on some things than the same task in Windows, you want proof? check out specview. Under windows, Discreet's Lustre can play 2k real-time, can't do it under Linux. I can't wait for somebody to get there hands on these new G5's for some REAL benchmarks
for the people who claim that the G5 is behind the opteron or the P4, i will like to hear how is it possible.
The G5 top at 2 ghz, like the opteron supposed to top (but did not suceed it for the moment) it's build upon SOI ,013 8 layers (best fab process of the moment) , has 58 millions transistors, 12 process units, the best bus of the PC market, 225 instructions in fligh , until 5 operations per cycle (compared to 3 for the 7455).
X 86 processor are a waste of transistors, because they need a huge decode unit at the contrary of the G5, and they did not have a SIMD unit as efficient as the G5.
The G5 is full SMP, at the contrary of the Xeon who shared the same bus.
The G5 is on the more modern mobo of the time, with 128 wide (and not dual channel) 400 mhz DDR ram, with hypertransport link, Sata disk, PCI-X slots.
IBM and Apple must suck terribely if with such features , they canno't compare with the X-86 world.
i went to the "idiots" website and read what he had to say.
he say hes a macuser but had nothing good to say about macusers nor the new powermacs.
this is good!
for a long time apple could not legitamately challenge the wintel boxes.
now it can and the wintel dummies will come for us,watch and see.
yesterday was a glorious day in mac folklore,but i belive its only the beginning.
long live apple!!!!!!
Yeah I thought that was telling also. Every so often and at the end he slips into this " Macuser's are fanatics " mode. Pretty strange for Macuser. Not so strange for someone trying to prove something that they've already made up their mind about.
wow my first post, but i think people look for too much in it all. as was stated earlier a few times no benchmarking, especially synthetic is good to judge to different computer products. hell, like nvidia who was engineering their drivers to score better in 3dmark tests. maybe it's the fastest pc in the world. maybe it's not. Depending on what you do youll find a reason for a computer to be best for the work. I know guys who program vhf radios who have to use an old pentium 90 rig. new machines won't communicate.
The G5 is faster tahn the G4, and has room to grow which I can't say for motorola as from my experience inother fields they suck ass.
wait till you see how much the system bus increases rendering and video processing.
They clearly published exactly what was used, and how. That they used products that may, MAY, have skewed the results in their favor is not a crime (Otherwise MSFT would be buried, burned).
The writer of the haxial blog seems to think that this is wrong. His opinion, and he can keep it (Exactly which bodily orifice I'd choose to plant it in is a different story).
I do wish that there were production G5s out there for people to test real-world performance.
I'm actually looking forward to a mag like PC world putting it through its paces. They've actually been giving macs a pretty fair shake lately.
Also, who gives a rat's ass about SPEC?
I mean, apple can rig things to a certain extent in the application tests I'm sure. But when the G5 smokes the competition THAT much, in THAT many programs, there's something to it.
Comments
It's pretty clear that the top end PPC now handily smokes the top end wintel, untill we see some Athlon64 numbers... Still, if I'm spending 2000-3000 on a photoshop or film or 3d animation rig, then the PM G5 gets my money. In the 1000-1500 range it's a harder call, and the PC offers more bang for the buck, but at 1299, the single 1.25Ghz G4 with L3 is not too shabby either.
Sounds like sour grapes. Apple (finally) has a no-nonsense chip with room to grow and some people's fragile worlds are collapsing.
http://www.amdzone.com/articleview.cfm?articleid=1296
http://www.amdzone.com/articleview.c...id=1296&page=2
SPECint rate and SPECfp rate scores that Apple provided for the G5 were 2P scores._ The P4 of course could not be tested in 2P mode, but the Xeon was._ Things look good for Apple's own SPEC benchmarks of the Xeon and P4 until you pan up at the official SPEC scores._ So which one is correct?_ The Apple scores or the ones on SPEC's site._ My money is with SPEC._ The difference is ridiculously large, and far too much for me to believe without some form of evidence._ There was no 1.7GHz Power4 to compare against in this test, but the 2P Opteron blows away the faster MHz G5.
So what is the true story of G5 performance?_ It certainly greatly improves on the G4, but I doubt we will ever find that out from Apple what true performance against modern PCs is. When they trotted out Adobe as a benchmark I tuned out._ Their ability to bend Photoshop benchmarks to suite their performance needs is legendary._ I wanted to hold out hope for Apple's PowerPC offering, but Jobs lies and stretches do a disservice to IBM and their powerful CPU._ This type of show from Apple may keep them pegged at 1.9% marketshare, but the sweet spot for real desktops is firmly below the$1,000 mark._ The PowerPC G5 system will allow Apple to hang onto what is left of their share, but they throw logic out the window with their over the top presentation that misleads, distorts facts, and leaves serious questions about actual performance._ Apple would do well to bring back clones and give Jobs the boot._ Here's hoping they do so they can regain marketshare and offer real competition to Microsoft, and further pressure on Intel._ Hand the real thanks to IBM and all of the borrowed PC technology._ That is what is allowing Apple to survive.
Those bench marks looks convincing to me, Apple Cheated on their specs and bechmark tests and I feel deceived by a cheat slight of hand dog and pony show carnival trick! (in the keynote demonstration)
-tom w
Originally posted by Matsu
It's the real world tests. Get mac app, get PC, face off.
It's pretty clear that the top end PPC now handily smokes the top end wintel, untill we see some Athlon64 numbers... Still, if I'm spending 2000-3000 on a photoshop or film or 3d animation rig, then the PM G5 gets my money. In the 1000-1500 range it's a harder call, and the PC offers more bang for the buck, but at 1299, the single 1.25Ghz G4 with L3 is not too shabby either.
[a little of topic again]
when i saw the prices of the G5 yesterday, i thought matsu would be pissed of and complaining about the $1999.00 pricetag of the low end. the 1.25Ghz G4 SP at $1299.00 would probally be $300.00 to much...
but that didn't happen.
so i will:
the lowend G5 is waaaaaaaaaaaaaaaaaaaaay to expensive, even the 1.8Ghz is a better option, no wait, the 2.0Ghz dp is the best deal...
oh well, wait the minute. back in august 2002, almost everybody who bought a powermac, bought the low end 867Mhz dp because that was the best bang for buck. now apple squeezed the specs of the low end G5 that much and give it a "confusing" pricetag of $1999.00. very close to the "better" option. in the netherlands the good/better difference is just 190 euroos. what would you buy? the better option i guess.
also this "high" pricetag of the low end gives lots of space for other products in the consumer matrix to find a comfortable place even though they're not (yet) upgraded to a G5
Originally posted by scotio
I do not know enough to comment on all the technical merits of this argument; I do however know, that reading the comments at the bottom of his critique made me, for the first time ever, thoroughly ashamed to be a Mac user.
No need. The numbers at the SPEC site are skewed, 'unfair.' It's the way things go.
Originally posted by Barto
It is disappointing that Apple *appears* to have cheated somewhat on the spec test. I would expect better from Apple than say, Gateway. Who compared Quake 3 on a Gateway to Quake 3 running in Classic on an iMac.
Barto
I dont understand why you feel that Apple cheated on the SPEC test. Not that I give a wilted patunia's rear about SPEC test or any other benchmark out there. I dont run benchmarks on my computer, but rather I do work on it. Apple took a $4K Dell system and kicked its rear with a $3K system. It doesnt get any simpler than that as far as I am concerned.
don't ever try to battle your enemy on it's own ground
if you use their tactics, even though you're better, you always lose because the place is not yours, go home.
use your own tactics on your own playing field and if you're better or not you win.
As Matsu says. Let's get the Apps and run'em. That'll tell you all you need to know. There are too many shady things going on for me to trust Spec %100
These are all just observations. In the end, I don't care about who is the fastest because PC's don't run mac OS X, and that was the real star yesterday.
Originally posted by Henriok
Intel uses their own especially for spec-tests tweakd compilers. Apple used the same compiler for every system. Apple is doing fair benchmarks.. the official spec-scores are quite misleading since every vendor supplies their own benchmarks, using whatever compiler they chose. AND.. they do not account for AltiVec.
Pixar uses Intel's especially for spec-tests tweakd compiler for Rendeman on x86, go figure... Face it, gcc is a slow compiler on both platforms, but Apples bigest crime in its number is useing tha NAGware compiler for the SPECfpu. Linux is alot slower on some things than the same task in Windows, you want proof? check out specview. Under windows, Discreet's Lustre can play 2k real-time, can't do it under Linux. I can't wait for somebody to get there hands on these new G5's for some REAL benchmarks
regards
a.
The G5 top at 2 ghz, like the opteron supposed to top (but did not suceed it for the moment) it's build upon SOI ,013 8 layers (best fab process of the moment) , has 58 millions transistors, 12 process units, the best bus of the PC market, 225 instructions in fligh , until 5 operations per cycle (compared to 3 for the 7455).
X 86 processor are a waste of transistors, because they need a huge decode unit at the contrary of the G5, and they did not have a SIMD unit as efficient as the G5.
The G5 is full SMP, at the contrary of the Xeon who shared the same bus.
The G5 is on the more modern mobo of the time, with 128 wide (and not dual channel) 400 mhz DDR ram, with hypertransport link, Sata disk, PCI-X slots.
IBM and Apple must suck terribely if with such features , they canno't compare with the X-86 world.
Originally posted by anthrax
I can't wait for somebody to get there hands on these new G5's for some REAL benchmarks
regards
a.
I agree with that!
-tom w
Originally posted by geekmeet
i went to the "idiots" website and read what he had to say.
he say hes a macuser but had nothing good to say about macusers nor the new powermacs.
this is good!
for a long time apple could not legitamately challenge the wintel boxes.
now it can and the wintel dummies will come for us,watch and see.
yesterday was a glorious day in mac folklore,but i belive its only the beginning.
long live apple!!!!!!
Yeah I thought that was telling also. Every so often and at the end he slips into this " Macuser's are fanatics " mode. Pretty strange for Macuser. Not so strange for someone trying to prove something that they've already made up their mind about.
The G5 is faster tahn the G4, and has room to grow which I can't say for motorola as from my experience inother fields they suck ass.
wait till you see how much the system bus increases rendering and video processing.
They clearly published exactly what was used, and how. That they used products that may, MAY, have skewed the results in their favor is not a crime (Otherwise MSFT would be buried, burned).
The writer of the haxial blog seems to think that this is wrong. His opinion, and he can keep it (Exactly which bodily orifice I'd choose to plant it in is a different story).
I do wish that there were production G5s out there for people to test real-world performance.
Also, who gives a rat's ass about SPEC?
I mean, apple can rig things to a certain extent in the application tests I'm sure. But when the G5 smokes the competition THAT much, in THAT many programs, there's something to it.