Apple Still Claims G5 Faster than Intel Xeon
"Does It Faster
The PowerPC G5 processor dramatically accelerates performance in real world applications. When compared head-to-head against PCs in a series of Photoshop tests, the dual 2.7GHz, 2.3GHz and 2.0GHz Power Mac G5 systems ran the 45 filters 98%, 78% and 59% faster, respectively, than the 3.6GHz Pentium 4-based system, and 72%, 56% and 38% faster than the dual 3.6GHz Xeon-based system.(2)"
-www.apple.com/powermac/ 9/17/05
I'm still not convinced any of Intel's Pentium or Xeon chips will have the performance of the G5's (except for maybe the Itanium 2 chips, which are poorly supported in software). When the dual-core G5's hit the market, Intel will be even further behind. I don't argue that Intel has a better power
erformance ratio, but for raw horsepower, they still have not leaped passed the G5. In fact they haven't even caught up to AMD with their Athlon 64 X2 4800+ (Dual Core 64-bit)
The PowerPC G5 processor dramatically accelerates performance in real world applications. When compared head-to-head against PCs in a series of Photoshop tests, the dual 2.7GHz, 2.3GHz and 2.0GHz Power Mac G5 systems ran the 45 filters 98%, 78% and 59% faster, respectively, than the 3.6GHz Pentium 4-based system, and 72%, 56% and 38% faster than the dual 3.6GHz Xeon-based system.(2)"
-www.apple.com/powermac/ 9/17/05
I'm still not convinced any of Intel's Pentium or Xeon chips will have the performance of the G5's (except for maybe the Itanium 2 chips, which are poorly supported in software). When the dual-core G5's hit the market, Intel will be even further behind. I don't argue that Intel has a better power
erformance ratio, but for raw horsepower, they still have not leaped passed the G5. In fact they haven't even caught up to AMD with their Athlon 64 X2 4800+ (Dual Core 64-bit)
Comments
Whether or not the G5 surpasses the Xeon at server-related tasks is a more clear subject for debate and analysis.
Originally posted by ngmapple
"Does It Faster
The PowerPC G5 processor dramatically accelerates performance in real world applications. When compared head-to-head against PCs in a series of Photoshop tests, the dual 2.7GHz, 2.3GHz and 2.0GHz Power Mac G5 systems ran the 45 filters 98%, 78% and 59% faster, respectively, than the 3.6GHz Pentium 4-based system, and 72%, 56% and 38% faster than the dual 3.6GHz Xeon-based system.(2)"
-www.apple.com/powermac/ 9/17/05
I'm still not convinced any of Intel's Pentium or Xeon chips will have the performance of the G5's (except for maybe the Itanium 2 chips, which are poorly supported in software). When the dual-core G5's hit the market, Intel will be even further behind. I don't argue that Intel has a better power
Yeah, that is like a 3 year old quote. New Dual core Xeons would kill the G5.
Originally posted by onlooker
Yeah, that is like a 3 year old quote. New Dual core Xeons would kill the G5.
Perhaps, but IBM now has dual core G5's. Once those make it into the PM (Paris?), I think the G5 will end up back on top.
When I first heard Apple on Intel announced I thought Intel would be making special non-x86 cpu's just for the Mac platform. Like high performance RISC cpu's based on Intel's XScale for instance. When I found out Apple was planning to use the same exact x86 cpu's featured in your average PC, I was dissapointed.
Those are marketing quotes. Entirely grain of salt material anyway.
Originally posted by nowayout11
Doesn't really matter what chips were more powerful 2 years ago, or what's even more powerful now. What matters is what's more powerful in 2 years when the switch is done.
Those are marketing quotes. Entirely grain of salt material anyway.
Couldn't agree more, but the information out there suggests Intel will be in two years only marginally past where AMD is today.
Its possible I am misunderstanding you so please correct me if I am wrong. but to me it looks like at current the current time a G5 is still faster at those tasks that they benchmark.
Originally posted by cubist
Guys, it really doesn't matter. Performance isn't the main consideration here, as THT points out in the Sossaman thread. Apple is aiming for 'good enough' performance in an inexpensive, attractive package. In other words, they are aiming for market share, not the high-performance niche.
If they don't come through with a high performance machine when they now have no excuses they will lose that niche completely. That niche is the home root of their user base. If they loose the dedicated all they have left is a bunch of people that could hardly give a crap, and they are all just a bunch of pliable followers anyway. The Apple faithful that stood with them this far will be gone, and Apple will have to compete on their own after that. It's going to be pretty lonely out there for Apple when they are considered the new Packard Bell of PC's once they are running on intel. " 'good enough' performance "
Originally posted by zenatek
I would just like to question why you guys are saying those quotes or scores are 2 or 3 years old? It states at the bottom they were done in april 2005 and also how would they have a 2.7dual G5 to test 2 or three years ago?
Its possible I am misunderstanding you so please correct me if I am wrong. but to me it looks like at current the current time a G5 is still faster at those tasks that they benchmark.
I'm not talking about chips two years old, we're talking about chips that Intel won't release until 2 years into the future. Comparing AMD 2005 to Intel 2006/2007. And AMD 2005 is performaning better than Intel 2006/2007 in benchmarks.
Originally posted by onlooker
It's going to be pretty lonely out there for Apple when they are considered the new Packard Bell of PC's once they are running on intel. " 'good enough' performance "
I couldn't agree more with you onlooker, Apple is a high performance brand first and foremost. Apple's PowerMac and PowerBook series are still their flag ship products. It has only been the last 5 years or so that Apple has entered the low cost (<$2000USD) market with it's iMac, iBook and mini product lines. These product lines are very important as they have helped gain a lot of market share. But that does not mean Apple as a whole should be satisfied with "good enough" performance, especially in it's Power lines.
Apple's long time customer base includes artists, video content creators, publishers, scientists, etc... These people look for the latest and most powerfull hardware IN ADDITION to a user experience like no other. For these users, it's about more than just the pretty enclosure and Aqua eye-candy. And additionally these customers are willing to pay the premium cost of "the best" hardware availible (I don't see anyone in this market segment trading a 64-bit machine for a 32-bit one to save a few bucks). Thus I do not expect Apple's Power hardware to become "as good as" the average PC, I still expect it to be better. Apple can still achieve that in a move to x86, but they need to do the following three tasks:
1- Use the most powerfull x86 cpu's availible (not cpu's that are "close" to being the most powerfull), while maintaining or improving upon battery life in the portables from current G4 battery life
2- Add in unique performance hardware features such as DSP or Xilinx/Altera FPGA co-processors for added performance (ie. to impliment updatable hardware codecs for faster encoding of video, audio, etc...)
3- Become an early adopter of non-performance related hardware (ie. Blueray DVD-R's, OLED screens, etc...)
I seem to remember that the inital feedback suggesting that the x86 based Mac trounced the dual 2.7GHz Power Mac G5?
Originally posted by ngmapple
Apple can still achieve that in a move to x86, but they need to do the following three tasks:
1- Use the most powerfull x86 cpu's availible (not cpu's that are "close" to being the most powerfull), while maintaining or improving upon battery life in the portables from current G4 battery life
Or use x86 cpus with a very high performance/watt ratio, and put lots of them on the motherboard. x86 cpus are usually cheaper than IBM PPCs, so that would not impact the final cost so much. And OS X has been tuned for SMP performance lately (although there is some extra work to do here!).
Dual dual-core Yonah in Powerbooks, single dual-core Yonah in iBooks, and quad dual-core Merom in Powermacs. That could beat AMD without having to do extra research on materials that don't melt at 300°C...
2- Add in unique performance hardware features such as DSP or Xilinx/Altera FPGA co-processors for added performance (ie. to impliment updatable hardware codecs for faster encoding of video, audio, etc...)
Apple (and 3rd party developers) would need to tweak huge parts of their code to take advantage of this. Such architectures are good for game consoles and embedded systems, but it is not for computers that need to efficiently run "legacy" code.
I believe we will see Apple relying more and more on the GPU in the future, instead of throwing money in conceiving coprocessors.
3- Become an early adopter of non-performance related hardware (ie. Blueray DVD-R's, OLED screens, etc...)
They definitely MUST do that to satisfy the pro niche market (and to satisfy the geeks we are).
Very nice post, anyway, ngmapple!
Originally posted by ngmapple
I couldn't agree more with you onlooker, Apple is a high performance brand first and foremost. Apple's PowerMac and PowerBook series are still their flag ship products. It has only been the last 5 years or so that Apple has entered the low cost (<$2000USD) market with it's iMac, iBook and mini product lines.
I guess you're just too young to remember the Plus, the Classic, the LC's, the IIsi, the Performas, and on and on and on. Apple's reputation has always come from the high end, but the backbone of its sales and its installed base has been on the low end (mid-end compared to the rest of the market) for a long time now. The Mac was born as a low-end computer ("for the rest of us"), after all.
Not surprisingly, Apple has historically stumbled worst when its low end has sagged. And when its fortunes rise, it's been driven from the low end - think LC and the early 90s boom, and the more recent iMac-driven resurgance. Compare to the Performa mess in the mid-90s. Today, the low end is flagging again, while the high end is in good shape. Steve said the Intel switch would start at the low end, and I think there's a good reason for that.
Originally posted by Messiah
I wonder if it'll end up in a situation where it's mid-range kit outperforms it's high-end kit ? like it did with the G3 and the 9600?
It certainly will... Except that this time, the high-end boxes will have software compatibility on their side. Remember the OS X transition, when all DTP and creative professionals were waiting for Quark to release an OS X friendly version of XPress before moving to OS X!
Originally posted by Towel
Steve said the Intel switch would start at the low end, and I think there's a good reason for that.
But when did he actually say that? In other threads we have all been going through that keynote, and nobody can seem to find anything he said about the low end coming first. Also, Apple will probably update their products as the suitable, or scheduled processors become available for each product. I doubt they'll wait for the low end processors to become available; if the high-end processors become available first, and their design is done for the pro machine they will more than likely release it. I really can't imagine them waiting to get this transition over for anything.
Originally posted by cubist
Apple is aiming for 'good enough' performance in an inexpensive, attractive package. In other words, they are aiming for market share, not the high-performance niche.
I'll believe it when I see it. Didn't Jobs' said he was going after the other 95% of the market 4 or 5 years ago? Sure, he may now have the luxury of lower priced chips from Intel, but I don't expect Apple to give up it's huge margins for marketshare especially in the pro segment.
By going Intel Apple leveled the playing field. Not they will never be too far behind (and never for a short blink of an eye be ahead). Speed has always been a crucial argument against Apple. They have taken that out of the equation and can now focus on a better OS, hardware/iPod integration (consumers) and exclusive video software (professionals).
The only ones who would actually be able to get an advantage by selecting Windows over Macs if they weren´t bleeding edge with the CPU is PS people. Everybody else would have other clear advantages. And if a 10% speed difference actually meant something for a PS artist she would already have anti-switched.