I would be very cautious about saying it "kicks ass" - better compilers will make it faster, but the same could be said of the P4. On the other hand, in order to go dual processor with Intel, you need Xeons.
We still haven't seen very much from AMD - hopefully they have the financial resources to remain competitive.
The G5 is a fine chip, and it at least gives us raw performance parity. If IBM can meet or exceed expectations, we may regain the absolute performance crown of winning on any benchmark. To that we can add the crown jewels of Altivec and one sexy OS.
Their software is dual processor aware, they just didn't graph the dual processor scores because its not a fair comparison against single processor machines. They do list their scores, however, and it demonstrates an almost doubling of performance, showing that the 970 is indeed very good at SMP.
They also note the AltiVec performance which is clock-for-clock equivalent to the G4 -- also impressive.
And they note that cheaper fast solutions may be had, but frankly if you want to buy a $500 machine go ahead. I've seen enough of them fail in all sorts of ways that I'm not prepared to waste my time using them. The price difference between the G5 and a super-cheap PC is gone very quickly if I have to spend much time at all struggling with the machine, and every minute I use the PC is a struggle. If you buy a quality PC then you're in the same price ballpark as Apple.
1) All of these tests were single-CPU tests. If they had a multi-threaded benchmark, or the benchmark was multi-processor aware, you could expect to see significant performance gains from the G5 specs
2) They are using an old Jet3D binary, one that was not compiled using gcc 3.3, and thus it doesn't have the 970-specific optimizations and scheduling in it. They do note this, but it is worth repeating here, because recompiled apps will benefit quite a bit in some cases from gcc 3.3 for the G5
Still, the scores look pretty good -- I am eager to see how the benchmarks turn out when they are recompiled under gcc 3.3, on shipping G5's (and running Panther, too )
I would be very cautious about saying it "kicks ass" - better compilers will make it faster, but the same could be said of the P4.
right, but the changes made to optimize code output and scheduling for the G5 are much more significant. With its shallow pipelines, we've been getting away with throwing just about any code at a G4 and having it work fairly well. The 970, like the P4, needs code that is fairly well optimized for it, to keep the deep pipelines full.
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
Jeezus. You haven't been banned yet? What's taking the Mods so long. Existence..this is Appleinsider not Battlefront.
Recent Existence contibutions to the board
Quote:
7-03-2003.What's worse is that Apple's claim of being the first 64-bit Desktop computert is also looking like utter tripe. It's verty likely Athlon64 desktops will ship before the G5.
Quote:
6-20-2003.Only 2GHz? It would take at least 2.3GHz to beat Intel's 3.2GHz P4 that will be introduced Monday. At 2GHz, we can expect SPEC scores to be on par with an old 3.06GHz P4 on a 533MHz FSB. I don't even think the 2GHz machine will match the current 3.0GHz canterwood for most things given OS X's overhead. These PowerMacs will certainly be more expensive than comparable PCs.
Quote:
6-13-2003.The current dual processor G4 machines Apple sells all used a shared bus topology. That means you have both G4's stuck sharing that 166MHz Mpx bus and 1.3GBps. No modern PC still uses shared bus topology. Intel Xeon's have independent busses to main memory for each processor.
With Hypertransport, Apple will finally catch up to modern technology. Still, don't expect any miracles in terms of beating Intel. There's a reason why Pixar went with Xeon's and didn't wait for the PPC970.
Idiot...this is APPLEinsider.com. Take your PC trash somewhere else. Reasonable comparisons are fine but your PC fanboyism is growing tiresome. Most of us here don't go crapping on PC boards with excessive Macintosh blather. There are plenty of PC based forums in which you can extol your Grand PC Knowledge.
Its so hard to actually compare the mac and PC cause there are so many variables that you have to take into consideration with rawr number benches like these.
The only thing that matters to me is REAL WORLD real app performance. Thats what I`m waiting for is all those websites out there to get a hold of these G5`s and run them against the latest and grreatest intel and AMD I`m really interested in seeing how a mostly unbiased test ranks these bad boys.
Its nice to see Apple has gotten back into the game. Like steve said this thing has legs. Also in the video they state that they have already started testing prototypes for the next generation chip.
Its getting so fast now that I personally don`t even need the power that the current machines provide but it sure is nice to brag to my PC friends for once. "Yeah the current apple has a dual 1GHz bus so what now"?
It should also be noted that the G5 was running against a Linux box ... if XP had been the OS, the G5 would have had a slaughter fest ... Throw Panther into the mix, and the P.C. setup would have looked like it was using a Pentium II ... Celeron ... [giddy squeal]
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
Build a dual xeon on dells website with 160 gig serial ata ( oh wait you CAN'T) and with a 1ghz bus (oh wait you CAN'T)
Its gonna be more expensive than the dual g5. The point has been made over and over again. Do you want the be told again that YOUR WRONG?
Go build urself that 2.66 ghz machine that you think is superior for 500 bucks. its not faster than a dual g5 and neither is a dual xeon setup which cost 4gs
When you can show me photoshop pumping stuff out faster than a g5 then you can rave all you want.
I think we wasting our time. Existence is an Appleinsider virus.
Using barebones components the cheapest I could find from a "reputable" retailer(hdnw.com) I found that $530 was as good as I could do for a 533Mhz bus P4 2.6Ghz. And this is really going cheapo with the Case...no sound card and other components. No thanks I'll stick to Macs.
The NASA reviewer clearly said he was using non optimized code for the G5. No changes were made. This wasn't meant to be a PC vs Mac test. It was a simple evaluation.
Oh and Xeons are NOT Point to Point busses. They are still shared in Dual Configs. Don't know where you got that from.
All these benchmarks are meaningless for two reasons. Firstly, they are all using out of date software. Second, Apple said world's fastest PC, not chip. The 1 ghz bus, PCI-X, FireWire 800, and 8 GIGS of fast RAM make this thing just plain dust a Windows machine at anywhere near the same price. It even dusts "workstations" with far higher prices.
Does the AMD still R0X0RS YOUR B0X3RS AND XP OWN3Z Y0UR S0UL?
Comments
The test software is NOT dual processor aware or G5 optimized, yet the dual 2.0 Power Mac still kicks ass!
There's gonna be a lot of "egg on my face" techies who have been bad mouthing the G5 methinks.
I would be very cautious about saying it "kicks ass" - better compilers will make it faster, but the same could be said of the P4. On the other hand, in order to go dual processor with Intel, you need Xeons.
We still haven't seen very much from AMD - hopefully they have the financial resources to remain competitive.
The G5 is a fine chip, and it at least gives us raw performance parity. If IBM can meet or exceed expectations, we may regain the absolute performance crown of winning on any benchmark. To that we can add the crown jewels of Altivec and one sexy OS.
Originally posted by MacsRGood4U
The G5 Powermac computer line will be released in late August. It is in the future. That's what this has to do with.
Does that mean all the G5 threads should go in here now?
http://arstechnica.infopop.net/OpenT...4&m=1210972375
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
Originally posted by Existence
Huh? Kicks ass? The obsolete 2.66GHz P4 beats the 2GHz G5. You can readily build or buy a 2.66GHz for <$500.
http://arstechnica.infopop.net/OpenT...4&m=1210972375
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
the 3.2GHz P4 will not beat the dual 2GHz G5...which teh P4 can't do, so you have to use the xeon which cost a bit more (i'm not sure how much though)
They also note the AltiVec performance which is clock-for-clock equivalent to the G4 -- also impressive.
And they note that cheaper fast solutions may be had, but frankly if you want to buy a $500 machine go ahead. I've seen enough of them fail in all sorts of ways that I'm not prepared to waste my time using them. The price difference between the G5 and a super-cheap PC is gone very quickly if I have to spend much time at all struggling with the machine, and every minute I use the PC is a struggle. If you buy a quality PC then you're in the same price ballpark as Apple.
2) They are using an old Jet3D binary, one that was not compiled using gcc 3.3, and thus it doesn't have the 970-specific optimizations and scheduling in it. They do note this, but it is worth repeating here, because recompiled apps will benefit quite a bit in some cases from gcc 3.3 for the G5
Still, the scores look pretty good -- I am eager to see how the benchmarks turn out when they are recompiled under gcc 3.3, on shipping G5's (and running Panther, too
Originally posted by dfryer
I would be very cautious about saying it "kicks ass" - better compilers will make it faster, but the same could be said of the P4.
right, but the changes made to optimize code output and scheduling for the G5 are much more significant. With its shallow pipelines, we've been getting away with throwing just about any code at a G4 and having it work fairly well. The 970, like the P4, needs code that is fairly well optimized for it, to keep the deep pipelines full.
Huh? Kicks ass? The obsolete 2.66GHz P4 beats the 2GHz G5. You can readily build or buy a 2.66GHz for <$500.
http://arstechnica.infopop.net/Open...24&m=1210972375
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
Jeezus. You haven't been banned yet? What's taking the Mods so long. Existence..this is Appleinsider not Battlefront.
Recent Existence contibutions to the board
7-03-2003.What's worse is that Apple's claim of being the first 64-bit Desktop computert is also looking like utter tripe. It's verty likely Athlon64 desktops will ship before the G5.
6-20-2003.Only 2GHz? It would take at least 2.3GHz to beat Intel's 3.2GHz P4 that will be introduced Monday. At 2GHz, we can expect SPEC scores to be on par with an old 3.06GHz P4 on a 533MHz FSB. I don't even think the 2GHz machine will match the current 3.0GHz canterwood for most things given OS X's overhead. These PowerMacs will certainly be more expensive than comparable PCs.
6-13-2003.The current dual processor G4 machines Apple sells all used a shared bus topology. That means you have both G4's stuck sharing that 166MHz Mpx bus and 1.3GBps. No modern PC still uses shared bus topology. Intel Xeon's have independent busses to main memory for each processor.
With Hypertransport, Apple will finally catch up to modern technology. Still, don't expect any miracles in terms of beating Intel. There's a reason why Pixar went with Xeon's and didn't wait for the PPC970.
Idiot...this is APPLEinsider.com. Take your PC trash somewhere else. Reasonable comparisons are fine but your PC fanboyism is growing tiresome. Most of us here don't go crapping on PC boards with excessive Macintosh blather. There are plenty of PC based forums in which you can extol your Grand PC Knowledge.
T
The only thing that matters to me is REAL WORLD real app performance. Thats what I`m waiting for is all those websites out there to get a hold of these G5`s and run them against the latest and grreatest intel and AMD I`m really interested in seeing how a mostly unbiased test ranks these bad boys.
Its nice to see Apple has gotten back into the game. Like steve said this thing has legs. Also in the video they state that they have already started testing prototypes for the next generation chip.
Its getting so fast now that I personally don`t even need the power that the current machines provide but it sure is nice to brag to my PC friends for once. "Yeah the current apple has a dual 1GHz bus so what now"?
Originally posted by Existence
Huh? Kicks ass? The obsolete 2.66GHz P4 beats the 2GHz G5. You can readily build or buy a 2.66GHz for <$500.
http://arstechnica.infopop.net/OpenT...4&m=1210972375
Oh, and the Portland FORTRAN compiler they use for the P4 is subpar.
Build a dual xeon on dells website with 160 gig serial ata ( oh wait you CAN'T) and with a 1ghz bus (oh wait you CAN'T)
Its gonna be more expensive than the dual g5. The point has been made over and over again. Do you want the be told again that YOUR WRONG?
Go build urself that 2.66 ghz machine that you think is superior for 500 bucks. its not faster than a dual g5 and neither is a dual xeon setup which cost 4gs
When you can show me photoshop pumping stuff out faster than a g5 then you can rave all you want.
I think we wasting our time. Existence is an Appleinsider virus.
Using barebones components the cheapest I could find from a "reputable" retailer(hdnw.com) I found that $530 was as good as I could do for a 533Mhz bus P4 2.6Ghz. And this is really going cheapo with the Case...no sound card and other components. No thanks I'll stick to Macs.
The NASA reviewer clearly said he was using non optimized code for the G5. No changes were made. This wasn't meant to be a PC vs Mac test. It was a simple evaluation.
Oh and Xeons are NOT Point to Point busses. They are still shared in Dual Configs. Don't know where you got that from.
Does the AMD still R0X0RS YOUR B0X3RS AND XP OWN3Z Y0UR S0UL?