windows media 10HIDEF is virtually the same specs of quality and compression as h.264. On windows this codec plays fine on even the lowly celeron. Even 3rd party h.264 hi-def content plays fine (not just wmv) either the machines I'm using (lowly laptops, take advantge of video card functions or windows is just more effient)
either way if apple plans to play hd on these new laptops they better come with a onboard chip for this (hopefully in way of new video cards) Apple is netorious for shipping cards with nice onboard hardware features and not taking advantage of them in laptops.
NOT 1080p. 720p might. But NOT 1080p. Sorry.
Just go to any of the other tech sites out there that have been following this.
This article tells you what the approx needs are. I posted it on another thread as well. A dual 533 G4 machine will play 720p as well, but there may be some dropped frames.
Rob Enderle, principal (and only) analyst for his company, the Enderle Group, is predicting and making statements out of his imagination again. By the way, when your company has one person in it, except for assistants and secretaries, call it a "Group".
we need a good mac vs windows bake-off site with documented, reproducible results to really get some good data on this stuff...
drips and drabs from apple, adobe, intel, barefeats, anandtech, etc, while useful, i think is just scratching the surface of the real-world benchmarks that we want to see.
apple commissioned a decent report on g5 and creative workflow but i think in that case it was a bit too technical, perhaps the general user/ creative user just wants to see, okay, adobe photoshop, choose 10 common tasks, compare mac vs pc setups.
now to find a venture capitalist a catchy domain name and access to lots of mac and windows pc hardware and software. and get a macintel developer box for skunkworks benchmarking to be revealed once real macintels start shipping
we need a good mac vs windows bake-off site with documented, reproducible results to really get some good data on this stuff...
drips and drabs from apple, adobe, intel, barefeats, anandtech, etc, while useful, i think is just scratching the surface of the real-world benchmarks that we want to see.
apple commissioned a decent report on g5 and creative workflow but i think in that case it was a bit too technical, perhaps the general user/ creative user just wants to see, okay, adobe photoshop, choose 10 common tasks, compare mac vs pc setups.
now to find a venture capitalist a catchy domain name and access to lots of mac and windows pc hardware and software. and get a macintel developer box for skunkworks benchmarking to be revealed once real macintels start shipping
Once Macintels ship there won't be any shortage of those. As testers can use Intel's compiler to do tests with we will find out if there is any appreciable OS difference between the boxes.
It will be interesting to see if Apple will be trying to get the top performance in each catagory or not. That's something we don't know yet. We can assume a lot, but it's worthless.
Once Macintels ship there won't be any shortage of those. As testers can use Intel's compiler to do tests with we will find out if there is any appreciable OS difference between the boxes.
It will be interesting to see if Apple will be trying to get the top performance in each catagory or not. That's something we don't know yet. We can assume a lot, but it's worthless.
well yes, once the hardware is standardised and assuming there are no problems with dual-booting windows xp and mac os x, that would be a good time to unleash the benchmarking frenzy.
for some of us can you clarify what would the Intel compiler thing be really used for? what sort of tests with the Intel compiler?
well yes, once the hardware is standardised and assuming there are no problems with dual-booting windows xp and mac os x, that would be a good time to unleash the benchmarking frenzy.
for some of us can you clarify what would the Intel compiler thing be really used for? what sort of tests with the Intel compiler?
Intel has a compiler which they came out with quite a while ago that they intended to be used mainly for testing. Some companies actually use it to compile commercial programs, but most programs are not compiled with it.
The benchmark tests used have to be compiled just like any other programs. Intel's compiler is a highly optimized compiler. It uses every method to squeeze the last flop of performance out of a chip.
The point about it of course, is that it is designed to get the last bit of performance out of INTEL'S chips. It don't work so good with AMD's chips, and of course, it won't work at all with the PPC.
When we see the PC crowd get all huffy about Apple's benchmarking, it's because Apple doesn't use benchmarks compiled with Intel's compiler. They use GCC on both machines. GCC has not been highly optimized for either platform, and is used to write more programs than Intel's is.
Therefore it's probably more fair to use that. But the PC crowd isn't happy about it because it doesn't show the x86 off in the best light.
IBM also has a highly optimized compiler for the PPC, but for whatever reason Apple (or the company that does the tests for Apple) has not chosen to use it.
I don't know at this time whether Intel's compiler can be used "out of the box" on a Macintel, or whether some changes will have to be made. Intel will probably be interested to get it to work on the Mac though.
Intel has a compiler which they came out with quite a while ago that they intended to be used mainly for testing. Some companies actually use it to compile commercial programs, but most programs are not compiled with it.
The benchmark tests used have to be compiled just like any other programs. Intel's compiler is a highly optimized compiler. It uses every method to squeeze the last flop of performance out of a chip.
The point about it of course, is that it is designed to get the last bit of performance out of INTEL'S chips. It don't work so good with AMD's chips, and of course, it won't work at all with the PPC.
When we see the PC crowd get all huffy about Apple's benchmarking, it's because Apple doesn't use benchmarks compiled with Intel's compiler. They use GCC on both machines. GCC has not been highly optimized for either platform, and is used to write more programs than Intel's is.
Therefore it's probably more fair to use that. But the PC crowd isn't happy about it because it doesn't show the x86 off in the best light.
IBM also has a highly optimized compiler for the PPC, but for whatever reason Apple (or the company that does the tests for Apple) has not chosen to use it.
I don't know at this time whether Intel's compiler can be used "out of the box" on a Macintel, or whether some changes will have to be made. Intel will probably be interested to get it to work on the Mac though.
cool. but that's the weird thing here. if we're going to talk about real-world benchmarks, one would use obviously pre-compiled cross-platform apps such as adobe photoshop and lightwave 3d.
i don't see the pc world get all messsed up about their pcmark, 3dmark, and si software's sandra, which obviously doesn't have different versions for intel/amd/nvidia/ati...
personally i would be seriously pissed off if any real world benchmark testing gets hijacked by people using specialised apps compiled using intel to prove some gains over apple. (important edit: OR VICE VERSA)
broadly-used, real-world, shipping binaries, cross-platform apps, should be the focus here.
no common end-user downloads source code and compiles it. way too complex still. for most apps, utilities and games for end-users, you just download the binary.
by all means i could go to anyone that says "Oh safari on tiger is 10.4.0 soooo slow", get the cvs of webtoolkit and compile a fresh spanking new safari and say, "look, it is teh snaptastic" but all my "real-world" arguments go out the window at that point.
cool. but that's the weird thing here. if we're going to talk about real-world benchmarks, one would use obviously pre-compiled cross-platform apps such as adobe photoshop and lightwave 3d.
i don't see the pc world get all messsed up about their pcmark, 3dmark, and si software's sandra, which obviously doesn't have different versions for intel/amd/nvidia/ati...
personally i would be seriously pissed off if any real world benchmark testing gets hijacked by people using specialised apps compiled using intel to prove some gains over apple. (important edit: OR VICE VERSA)
broadly-used, real-world, shipping binaries, cross-platform apps, should be the focus here.
no common end-user downloads source code and compiles it. way too complex still. for most apps, utilities and games for end-users, you just download the binary.
by all means i could go to anyone that says "Oh safari on tiger is 10.4.0 soooo slow", get the cvs of webtoolkit and compile a fresh spanking new safari and say, "look, it is teh snaptastic" but all my "real-world" arguments go out the window at that point.
So. Who ever said that benchmarking has anything to do with "real-world" computing?
These benchmarks (BMK's) are there to show who has the most integers, flops, etc. It doesn't have much to do with memory bandwidth, HD performance etc. It doesn't have anything to do with game performance, graphics performance, or how fast Word scrolls down a larger document.
It's all about HORSEPOWER baby! My numbers are bigger than YOURS!
Actually, if you go to Anands, you will find games tested extensively with video cards etc. Other sites do this as well. CDFreaks does very well at testing CD and DVD drives and disks.
But almost all of this is PC stuff, there are fewer Mac testing sites, and frankly the Mac BMK programs, for the most part, suck. Maybe that's because we aren't as interested in this stuff as they are.
If you look at the PC enthusiast sites, you will see that they are mostly interested in how they can get that extra 2% of performance, even if it shortens the life of their machines by 25%. Since we can't tweak our machines that way, we don't try.
ALL Benchmark programs are useless apart from pissing contests.
Intel's compiler does very well in benchmarks because they've spent a lot of time optimizing it to get really good results in benchmark tests. In general use it's not as good.
IBM's xlc compiler however is quite good as a general purpose compiler. The problem being, it costs money so Apple wouldn't be able to ship it free with XTools like they do GCC.
XBench may not be a brilliant program as benchmarking goes but it does have one thing going for it - it's a universal binary. At the moment it's showing a huge disparity between PPC and Intel at running the same code for whatever reason be it OS or hardware.
However, that's beside the point. What matters is the performance of off-the-shelf applications as you can't recompile those and those are the apps we run everyday. If it turns out Photoshop on MacOSX Intel is slower than on Windows on the same kit, Apple will have serious questions to answer about their compiler of choice.
If it's slower than on MacOSX PPC, then they've another set of questions too.
It's always been a fudge comparing Windows and MacOSX for speed at running the same applications as you could often blame the hardware on each platform for any difference. Now Apple only have the software to blame as we can do direct comparisons on the same hardware.
However, that's beside the point. What matters is the performance of off-the-shelf applications as you can't recompile those and those are the apps we run everyday. If it turns out Photoshop on MacOSX Intel is slower than on Windows on the same kit, Apple will have serious questions to answer about their compiler of choice.
If it's slower than on MacOSX PPC, then they've another set of questions too.
It's always been a fudge comparing Windows and MacOSX for speed at running the same applications as you could often blame the hardware on each platform for any difference. Now Apple only have the software to blame as we can do direct comparisons on the same hardware.
cool. i like it when people agree with me
remember also that when comparing pre-compiled photoshop that it is a pandora's box in it's on way. commonly people set up the photoshop 'timedemo' as a script of actions eg. in the past several years steve showing how you put together a nice-looking bigass Pixar/other movie poster.
now that leaves it open to a whole new set of debates -- what was the workflow steps? what were the filters used? did they specifically use more OsX - favourable filters in the workflow? how representative are these steps of day-to-day photoshop use? for what industry? okay, now what about a Flash timedemo? web browsing? powerpoint?
i have a small research background in molecular biology somewhere in my past, so somehow i've picked up a habit of looking into research and nitpicking it to death...
but like aegis and melgross said, at least with the hardware variables eliminated with Macintels, anyone with a scientific inclination has gotta love that. one bigass parameter controlled, let the bloodletting really begin now.
btw, i love xBench when comparing between mac models**, but when it comes to macintel, i have no rationale yet but i'm not too concerned about it's prototype macintel results, even if it IS a universal binary.
**xBench was for example extremely crucial in highlighting the SATA 7200rpm hard disk "bottleneck issue" in revA iMac g5s which was subsequently addressed in revB iMac g5s
....also after a painful mod of a 5400rpm drive into my iBook g4 xBench helped me feel good that it was worth it. some numbers to back up my general impression that things were snappiertastic. heh
I can't see any particular reason why OS X software would be slower on CPU benchmarks than Windows software. If a company is building with the Intel compiler on Windows, they'll probably do so on MacOS as well. Also, GCC is apparently much better on x86 - within 20% of the Intel compiler according to some testing I've seen - and probably just as fast as the widely used MS compiler.
The areas of concern would be:
1) Software that still uses QuickDraw, because apparently that's considerably slower than the new APIs. OTOH, Windows GDI will probably be running in a similar penalty mode under Vista.
2) Games, particularly DirectX games that run through an OpenGL translation layer. Also the OS X Doom3 benchmarks supposedly pointed OpenGL drivers out a weak point. However, nobody buys Macs for the FPS, so no biggie.
3) Server benchmarks, such as I/O or multiprocessing, where OS X hasn't always shown well.
If a company is building with the Intel compiler on Windows, they'll probably do so on MacOS as well.
Except that the Intel Compiler is designed for the x86 (and possibly Itanium) architecture, and as you may or may not recall, Macs currently ship with PowerPC chips.
I can't see any particular reason why OS X software would be slower on CPU benchmarks than Windows software. If a company is building with the Intel compiler on Windows, they'll probably do so on MacOS as well. Also, GCC is apparently much better on x86 - within 20% of the Intel compiler according to some testing I've seen - and probably just as fast as the widely used MS compiler.
The areas of concern would be:
1) Software that still uses QuickDraw, because apparently that's considerably slower than the new APIs. OTOH, Windows GDI will probably be running in a similar penalty mode under Vista.
2) Games, particularly DirectX games that run through an OpenGL translation layer. Also the OS X Doom3 benchmarks supposedly pointed OpenGL drivers out a weak point. However, nobody buys Macs for the FPS, so no biggie.
3) Server benchmarks, such as I/O or multiprocessing, where OS X hasn't always shown well.
20% is considered to be a huge difference. The Mac gets smacked down if the diff is 10%
As I've already said, very few programs use Intel's compiler. Benchmarking programs sometimes do. It's Intel's big gun. You know, (alright, so you don't know) in the '80's it wasn't considered fair by Intel to use a benchmark that used all the features of another chip if Intel's chips didn't have those features. So when other chips has a built-in floating point unit, as the 68000 chips did, the software did the comparisons WITHOUT using it. How's that for fair?
This is what happens with benchmarking.
A couple of years ago Nvidia was caught cheating as well. When their boards detected benchmarking tests they changed the way they worked to make it SEEM as though they were faster. ATI was accused of something less serious, but similiar.
X Bench is infamous for never giving the same scores twice. I've run tests with it that ran a 40% dif. I'm not alone in this.
Except that the Intel Compiler is designed for the x86 (and possibly Itanium) architecture, and as you may or may not recall, Macs currently ship with PowerPC chips.
Gene, that's what we're talking about; the Macintel's.
Freed from the bounds of the G4 CPU, I don't believe 10%-20% is going to affect many Mac user's purchasing decisions.
Quote:
Originally posted by melgross
This is what happens with benchmarking.
Speaking of questionable benchmarketing, I'd rank the Apple report that claimed Dell's SPECMark scores were something like half the official numbers pretty highly. Apple didn't even have the nuts to officially submit their own numbers. (And unlike 3DMark, SPEC is usually a respected vendor-independant suite.)
Freed from the bounds of the G4 CPU, I don't believe 10%-20% is going to affect many Mac user's purchasing decisions.
Speaking of questionable benchmarketing, I'd rank the Apple report that claimed Dell's SPECMark scores were something like half the official numbers pretty highly. Apple didn't even have the nuts to officially submit their own numbers. (And unlike 3DMark, SPEC is usually a respected vendor-independant suite.)
I'm not quite sure I understand what you mean by the first part of the first sentence. Of course it's not only the G4.
Mac users may not cringe at a difference, but a switcher might. Apple is picking up marketshare from those who have never purchased a computer, but mostly from those who now use Windows. The comparison is nebulous when comparing between OS's on different platforms. It's much less so if both machines are using the same chips.
IBM has SPECmarks submitted. For PC's they are supposed to measure just the cpu, so it shouldn't matter that much.
Apple Computer is believed to be prepping one final update to its PowerPC-based PowerBook G4 product line that could be unveiled at the end of next month, AppleInsider has learned...
Two new models that have shown up on radar are referenced as the PowerBook5,8 and the PowerBook5,9. Both Apple's current 15-inch and 17-inch PowerBook G4 systems identify themselves as the PowerBook5,7, with the 12-inch model being listed as a member of the iBook family (PowerBook6,8).
Whatever happened to these references to two new PowerBooks?
Whatever happened to these references to two new PowerBooks?
If they are intended to use the new 7448 cpu's from Freescale, then that could be why we haven't seen them.
Apple's fortunes have waxed and waned depending on the ability of their suppliers to deliver what they have promised on time and at speed. That's why Apple is leaving them after all.
These chips are delayed. That could be why Jobs canceled his keynote at the Paris fair.
Comments
Originally posted by webmail
windows media 10HIDEF is virtually the same specs of quality and compression as h.264. On windows this codec plays fine on even the lowly celeron. Even 3rd party h.264 hi-def content plays fine (not just wmv) either the machines I'm using (lowly laptops, take advantge of video card functions or windows is just more effient)
either way if apple plans to play hd on these new laptops they better come with a onboard chip for this (hopefully in way of new video cards) Apple is netorious for shipping cards with nice onboard hardware features and not taking advantage of them in laptops.
NOT 1080p. 720p might. But NOT 1080p. Sorry.
Just go to any of the other tech sites out there that have been following this.
Otherwise, you are a unique person.
http://www.theinquirer.net/?article=25057
Originally posted by melgross
Rob Enderle, principal (and only) analyst for his company, the Enderle Group, is predicting and making statements out of his imagination again. By the way, when your company has one person in it, except for assistants and secretaries, call it a "Group".
It must be all those voices in his head.
drips and drabs from apple, adobe, intel, barefeats, anandtech, etc, while useful, i think is just scratching the surface of the real-world benchmarks that we want to see.
apple commissioned a decent report on g5 and creative workflow but i think in that case it was a bit too technical, perhaps the general user/ creative user just wants to see, okay, adobe photoshop, choose 10 common tasks, compare mac vs pc setups.
now to find a venture capitalist a catchy domain name and access to lots of mac and windows pc hardware and software. and get a macintel developer box for skunkworks benchmarking to be revealed once real macintels start shipping
Originally posted by sunilraman
we need a good mac vs windows bake-off site with documented, reproducible results to really get some good data on this stuff...
drips and drabs from apple, adobe, intel, barefeats, anandtech, etc, while useful, i think is just scratching the surface of the real-world benchmarks that we want to see.
apple commissioned a decent report on g5 and creative workflow but i think in that case it was a bit too technical, perhaps the general user/ creative user just wants to see, okay, adobe photoshop, choose 10 common tasks, compare mac vs pc setups.
now to find a venture capitalist a catchy domain name and access to lots of mac and windows pc hardware and software. and get a macintel developer box for skunkworks benchmarking to be revealed once real macintels start shipping
Once Macintels ship there won't be any shortage of those. As testers can use Intel's compiler to do tests with we will find out if there is any appreciable OS difference between the boxes.
It will be interesting to see if Apple will be trying to get the top performance in each catagory or not. That's something we don't know yet. We can assume a lot, but it's worthless.
Originally posted by melgross
Once Macintels ship there won't be any shortage of those. As testers can use Intel's compiler to do tests with we will find out if there is any appreciable OS difference between the boxes.
It will be interesting to see if Apple will be trying to get the top performance in each catagory or not. That's something we don't know yet. We can assume a lot, but it's worthless.
well yes, once the hardware is standardised and assuming there are no problems with dual-booting windows xp and mac os x, that would be a good time to unleash the benchmarking frenzy.
for some of us can you clarify what would the Intel compiler thing be really used for? what sort of tests with the Intel compiler?
Originally posted by sunilraman
well yes, once the hardware is standardised and assuming there are no problems with dual-booting windows xp and mac os x, that would be a good time to unleash the benchmarking frenzy.
for some of us can you clarify what would the Intel compiler thing be really used for? what sort of tests with the Intel compiler?
Intel has a compiler which they came out with quite a while ago that they intended to be used mainly for testing. Some companies actually use it to compile commercial programs, but most programs are not compiled with it.
The benchmark tests used have to be compiled just like any other programs. Intel's compiler is a highly optimized compiler. It uses every method to squeeze the last flop of performance out of a chip.
The point about it of course, is that it is designed to get the last bit of performance out of INTEL'S chips. It don't work so good with AMD's chips, and of course, it won't work at all with the PPC.
When we see the PC crowd get all huffy about Apple's benchmarking, it's because Apple doesn't use benchmarks compiled with Intel's compiler. They use GCC on both machines. GCC has not been highly optimized for either platform, and is used to write more programs than Intel's is.
Therefore it's probably more fair to use that. But the PC crowd isn't happy about it because it doesn't show the x86 off in the best light.
IBM also has a highly optimized compiler for the PPC, but for whatever reason Apple (or the company that does the tests for Apple) has not chosen to use it.
I don't know at this time whether Intel's compiler can be used "out of the box" on a Macintel, or whether some changes will have to be made. Intel will probably be interested to get it to work on the Mac though.
Originally posted by melgross
Intel has a compiler which they came out with quite a while ago that they intended to be used mainly for testing. Some companies actually use it to compile commercial programs, but most programs are not compiled with it.
The benchmark tests used have to be compiled just like any other programs. Intel's compiler is a highly optimized compiler. It uses every method to squeeze the last flop of performance out of a chip.
The point about it of course, is that it is designed to get the last bit of performance out of INTEL'S chips. It don't work so good with AMD's chips, and of course, it won't work at all with the PPC.
When we see the PC crowd get all huffy about Apple's benchmarking, it's because Apple doesn't use benchmarks compiled with Intel's compiler. They use GCC on both machines. GCC has not been highly optimized for either platform, and is used to write more programs than Intel's is.
Therefore it's probably more fair to use that. But the PC crowd isn't happy about it because it doesn't show the x86 off in the best light.
IBM also has a highly optimized compiler for the PPC, but for whatever reason Apple (or the company that does the tests for Apple) has not chosen to use it.
I don't know at this time whether Intel's compiler can be used "out of the box" on a Macintel, or whether some changes will have to be made. Intel will probably be interested to get it to work on the Mac though.
cool. but that's the weird thing here. if we're going to talk about real-world benchmarks, one would use obviously pre-compiled cross-platform apps such as adobe photoshop and lightwave 3d.
i don't see the pc world get all messsed up about their pcmark, 3dmark, and si software's sandra, which obviously doesn't have different versions for intel/amd/nvidia/ati...
personally i would be seriously pissed off if any real world benchmark testing gets hijacked by people using specialised apps compiled using intel to prove some gains over apple. (important edit: OR VICE VERSA)
broadly-used, real-world, shipping binaries, cross-platform apps, should be the focus here.
no common end-user downloads source code and compiles it. way too complex still. for most apps, utilities and games for end-users, you just download the binary.
by all means i could go to anyone that says "Oh safari on tiger is 10.4.0 soooo slow", get the cvs of webtoolkit and compile a fresh spanking new safari and say, "look, it is teh snaptastic" but all my "real-world" arguments go out the window at that point.
Originally posted by sunilraman
cool. but that's the weird thing here. if we're going to talk about real-world benchmarks, one would use obviously pre-compiled cross-platform apps such as adobe photoshop and lightwave 3d.
i don't see the pc world get all messsed up about their pcmark, 3dmark, and si software's sandra, which obviously doesn't have different versions for intel/amd/nvidia/ati...
personally i would be seriously pissed off if any real world benchmark testing gets hijacked by people using specialised apps compiled using intel to prove some gains over apple. (important edit: OR VICE VERSA)
broadly-used, real-world, shipping binaries, cross-platform apps, should be the focus here.
no common end-user downloads source code and compiles it. way too complex still. for most apps, utilities and games for end-users, you just download the binary.
by all means i could go to anyone that says "Oh safari on tiger is 10.4.0 soooo slow", get the cvs of webtoolkit and compile a fresh spanking new safari and say, "look, it is teh snaptastic" but all my "real-world" arguments go out the window at that point.
So. Who ever said that benchmarking has anything to do with "real-world" computing?
These benchmarks (BMK's) are there to show who has the most integers, flops, etc. It doesn't have much to do with memory bandwidth, HD performance etc. It doesn't have anything to do with game performance, graphics performance, or how fast Word scrolls down a larger document.
It's all about HORSEPOWER baby! My numbers are bigger than YOURS!
Actually, if you go to Anands, you will find games tested extensively with video cards etc. Other sites do this as well. CDFreaks does very well at testing CD and DVD drives and disks.
But almost all of this is PC stuff, there are fewer Mac testing sites, and frankly the Mac BMK programs, for the most part, suck. Maybe that's because we aren't as interested in this stuff as they are.
If you look at the PC enthusiast sites, you will see that they are mostly interested in how they can get that extra 2% of performance, even if it shortens the life of their machines by 25%. Since we can't tweak our machines that way, we don't try.
Intel's compiler does very well in benchmarks because they've spent a lot of time optimizing it to get really good results in benchmark tests. In general use it's not as good.
IBM's xlc compiler however is quite good as a general purpose compiler. The problem being, it costs money so Apple wouldn't be able to ship it free with XTools like they do GCC.
XBench may not be a brilliant program as benchmarking goes but it does have one thing going for it - it's a universal binary. At the moment it's showing a huge disparity between PPC and Intel at running the same code for whatever reason be it OS or hardware.
However, that's beside the point. What matters is the performance of off-the-shelf applications as you can't recompile those and those are the apps we run everyday. If it turns out Photoshop on MacOSX Intel is slower than on Windows on the same kit, Apple will have serious questions to answer about their compiler of choice.
If it's slower than on MacOSX PPC, then they've another set of questions too.
It's always been a fudge comparing Windows and MacOSX for speed at running the same applications as you could often blame the hardware on each platform for any difference. Now Apple only have the software to blame as we can do direct comparisons on the same hardware.
Originally posted by aegisdesign
......
However, that's beside the point. What matters is the performance of off-the-shelf applications as you can't recompile those and those are the apps we run everyday. If it turns out Photoshop on MacOSX Intel is slower than on Windows on the same kit, Apple will have serious questions to answer about their compiler of choice.
If it's slower than on MacOSX PPC, then they've another set of questions too.
It's always been a fudge comparing Windows and MacOSX for speed at running the same applications as you could often blame the hardware on each platform for any difference. Now Apple only have the software to blame as we can do direct comparisons on the same hardware.
cool. i like it when people agree with me
remember also that when comparing pre-compiled photoshop that it is a pandora's box in it's on way. commonly people set up the photoshop 'timedemo' as a script of actions eg. in the past several years steve showing how you put together a nice-looking bigass Pixar/other movie poster.
now that leaves it open to a whole new set of debates -- what was the workflow steps? what were the filters used? did they specifically use more OsX - favourable filters in the workflow? how representative are these steps of day-to-day photoshop use? for what industry? okay, now what about a Flash timedemo? web browsing? powerpoint?
i have a small research background in molecular biology somewhere in my past, so somehow i've picked up a habit of looking into research and nitpicking it to death...
but like aegis and melgross said, at least with the hardware variables eliminated with Macintels, anyone with a scientific inclination has gotta love that. one bigass parameter controlled, let the bloodletting really begin now.
btw, i love xBench when comparing between mac models**, but when it comes to macintel, i have no rationale yet but i'm not too concerned about it's prototype macintel results, even if it IS a universal binary.
**xBench was for example extremely crucial in highlighting the SATA 7200rpm hard disk "bottleneck issue" in revA iMac g5s which was subsequently addressed in revB iMac g5s
....also after a painful mod of a 5400rpm drive into my iBook g4 xBench helped me feel good that it was worth it. some numbers to back up my general impression that things were snappiertastic. heh
The areas of concern would be:
1) Software that still uses QuickDraw, because apparently that's considerably slower than the new APIs. OTOH, Windows GDI will probably be running in a similar penalty mode under Vista.
2) Games, particularly DirectX games that run through an OpenGL translation layer. Also the OS X Doom3 benchmarks supposedly pointed OpenGL drivers out a weak point. However, nobody buys Macs for the FPS, so no biggie.
3) Server benchmarks, such as I/O or multiprocessing, where OS X hasn't always shown well.
If a company is building with the Intel compiler on Windows, they'll probably do so on MacOS as well.
Except that the Intel Compiler is designed for the x86 (and possibly Itanium) architecture, and as you may or may not recall, Macs currently ship with PowerPC chips.
Originally posted by IntlHarvester
I can't see any particular reason why OS X software would be slower on CPU benchmarks than Windows software. If a company is building with the Intel compiler on Windows, they'll probably do so on MacOS as well. Also, GCC is apparently much better on x86 - within 20% of the Intel compiler according to some testing I've seen - and probably just as fast as the widely used MS compiler.
The areas of concern would be:
1) Software that still uses QuickDraw, because apparently that's considerably slower than the new APIs. OTOH, Windows GDI will probably be running in a similar penalty mode under Vista.
2) Games, particularly DirectX games that run through an OpenGL translation layer. Also the OS X Doom3 benchmarks supposedly pointed OpenGL drivers out a weak point. However, nobody buys Macs for the FPS, so no biggie.
3) Server benchmarks, such as I/O or multiprocessing, where OS X hasn't always shown well.
20% is considered to be a huge difference. The Mac gets smacked down if the diff is 10%
As I've already said, very few programs use Intel's compiler. Benchmarking programs sometimes do. It's Intel's big gun. You know, (alright, so you don't know) in the '80's it wasn't considered fair by Intel to use a benchmark that used all the features of another chip if Intel's chips didn't have those features. So when other chips has a built-in floating point unit, as the 68000 chips did, the software did the comparisons WITHOUT using it. How's that for fair?
This is what happens with benchmarking.
A couple of years ago Nvidia was caught cheating as well. When their boards detected benchmarking tests they changed the way they worked to make it SEEM as though they were faster. ATI was accused of something less serious, but similiar.
X Bench is infamous for never giving the same scores twice. I've run tests with it that ran a 40% dif. I'm not alone in this.
Originally posted by Gene Clean
Except that the Intel Compiler is designed for the x86 (and possibly Itanium) architecture, and as you may or may not recall, Macs currently ship with PowerPC chips.
Gene, that's what we're talking about; the Macintel's.
Originally posted by melgross
Gene, that's what we're talking about; the Macintel's.
Yeah, I kind of noticed that. Though to me it seems he's talking in present tense, so that means he was talking about current Macs.
If not, mea culpa.
Originally posted by melgross
This is what happens with benchmarking.
Speaking of questionable benchmarketing, I'd rank the Apple report that claimed Dell's SPECMark scores were something like half the official numbers pretty highly. Apple didn't even have the nuts to officially submit their own numbers. (And unlike 3DMark, SPEC is usually a respected vendor-independant suite.)
Originally posted by IntlHarvester
Freed from the bounds of the G4 CPU, I don't believe 10%-20% is going to affect many Mac user's purchasing decisions.
Speaking of questionable benchmarketing, I'd rank the Apple report that claimed Dell's SPECMark scores were something like half the official numbers pretty highly. Apple didn't even have the nuts to officially submit their own numbers. (And unlike 3DMark, SPEC is usually a respected vendor-independant suite.)
I'm not quite sure I understand what you mean by the first part of the first sentence. Of course it's not only the G4.
Mac users may not cringe at a difference, but a switcher might. Apple is picking up marketshare from those who have never purchased a computer, but mostly from those who now use Windows. The comparison is nebulous when comparing between OS's on different platforms. It's much less so if both machines are using the same chips.
IBM has SPECmarks submitted. For PC's they are supposed to measure just the cpu, so it shouldn't matter that much.
Originally posted by AppleInsider
Apple Computer is believed to be prepping one final update to its PowerPC-based PowerBook G4 product line that could be unveiled at the end of next month, AppleInsider has learned...
Two new models that have shown up on radar are referenced as the PowerBook5,8 and the PowerBook5,9. Both Apple's current 15-inch and 17-inch PowerBook G4 systems identify themselves as the PowerBook5,7, with the 12-inch model being listed as a member of the iBook family (PowerBook6,8).
Whatever happened to these references to two new PowerBooks?
Originally posted by ka2357
Whatever happened to these references to two new PowerBooks?
If they are intended to use the new 7448 cpu's from Freescale, then that could be why we haven't seen them.
Apple's fortunes have waxed and waned depending on the ability of their suppliers to deliver what they have promised on time and at speed. That's why Apple is leaving them after all.
These chips are delayed. That could be why Jobs canceled his keynote at the Paris fair.