I would like to point out that some of the tests on both systems are done with Quicktime and iTunes, which are really testing how well Apple ported the applications to Windows than the actual speed (and the Windows ports are definitely worse in terms of performance).
@Mio: EXACTLY! When I recently benchmarked my Core i7 Hackintosh against some other computers, I looked no farther than FFMPEG. Multi-threaded, multi-platform -- I could easily compare Apples to Oranges, erm, Windows to Ubuntu.
Actually, there are a number of items that could sway the numbers
First, Win7 64bit is a pure 64bit OS. SL requires you to hold down the 6 and 4 keys to boot as 64bit. Was this done? The article does not say.
For Windows to run, Apple has provided drivers. Are these Apple written drivers optimized for Win7 64bit or are they the older Vista drivers? Drivers that are a few years old compared to drivers that had optimizations performed over the years would have an impact. I could say that Apple may have written their Win drivers to perform slower, but I don't think (or I hope) they are not that evil. This would impact startup performance since those drivers need to load during startup, and impact shutdown since the drivers need to be closed, and have a performance hit over the time use of the computer.
They used QTX for SL, which the article states "Apple claims to have much improved performance over the previous version." Since QTX is not available for Windows, the numbers are going to go in favor of OSX. If you are using an optimized newer version vs. an older slower performing version, then you are going to have the results skewed towards OSX. They should have used software such as Zune 4.0, which not only is the newest version, but also is a 64bit app throughout. Since QTX is 64bit onSL and QT for Win7 is 32bit, you are going to have different results as encoding/decoding can take advantage of 64 bits, but can't when using QT on Win7.
Of course, this are real-world use so you need to take the numbers as they are if you were using a Mac as your hardware, but it would be interesting if you compared neutral software, or compared WMP/Zune vs. iTunes for those numbers (WMP would be less desirable as it is a 32 bit app).
There isn't a way to make this comparison valid, really. Apple builds the OS and the machine to work well together. Microsoft builds the OS to work on a huge variety of machine configurations. Since Microsoft doesn't have retail machines that Windows 7 was specifically designed for, the comparison falls apart, even if you put both OS'es on a third-party box.
I agree with you for the most part, but wouldn't multi-platform apps such as ffmpeg or vlc be more of fair comparison? There is a lot more to learn from fair benchmarks, like how the Macbook Pro got the highest Vista system rating in 2008.
I might add that starting a Mac in 64 bit mode only affects the kernal, the OS and apps still run in 64 bit even when kernal is in 32 bit mode.
@Mio Yeah complete bullshit, for consistency lets look at this: Different HDs even if they are spin at the same rate, have the same interface, and same amount of cache they are not the same HD. Brand new version of iTunes for SL and version 7.6.4 for Windows? Where is version 9 on Windows?
I have a more balanced comparison, how about you run your OSX/Win7 tests on the MBP and then you go get a Win7 based machine that costs the same amount as the MBP and compare that machines performance. Don't forget the pretty graphs.
@CtrlBurn - This is not exactly true. While they can pick and choose which hardware they use, for the most part all the hardware in a Mac is the same as a Windows computer. Really it comes down to the drivers in how the OS interacts with the hardware. The only place where the hardware has a real effect is that Mac uses EFI while Windows uses an older, slower BIOS. This will benefit SL and negatively impact Win7 as (if I remember correctly) BIOS is emulated when using Mac hardware.
@Tsing and BillG- Actually, if they boot as 64 bit it would benefit Mac as the kernel used is 32 bit unless you do the 6 and 4 key thing. Using a 32 bit kernel, which uses 32 bit drivers can, under some circumstances, adversely be a performance bottleneck compared to a 64 bit kernel with 64 bit drivers. For example, when they are reading from the CD/DVD drive, if they can read 64 bits at a time rather than reading 32 bits it would help perf. If the encoder needs to go to a driver or the kernel, there is a 64 to 32 bit thunk (using Windows terminology) and back which can be a perf bottleneck.
@BillG - which is why I wrote "it would be interesting if you compared neutral software"
Something else I just thought of that could skew the results (again, this it is fair since it is real world use, but an explanation of results), shouldn't the results be even better for iTunes on SL being they touted they supported CUDA (or whatever they were calling it), where encoding/decoding would give a great benefit? Or do they not use it in QTX?
Good lord, that NoHOne dude is so insecure about these results that he decided to write us a mini-novella to tell us about it. I would hate to see what he would write once they release the benchmarks for Macs with LightPeak.
Calm down, man. It's not the end of the world if Snow Leopard is a little faster. You'll live.
@Detox - you really are a sad person, aren't you? I am starting to believe, as mentioned yesterday, thay you are Paul being Paul wrote that same exact thing to me once, word for word.
If you were not such a pathetic Apple fanboy, you would see where I am saying things that would help SL have better numbers to bring them up to a more equal footing. You would also see where I am saying that there is no harm in using the most optimized software for each platform. If you want an exact performance comparison, then yes, you should use the exact same software with the only difference being the underlaying OS. But if you want to use the OS that gives you the best overall perf, then use iTunes with QTX vs. Zune 4.0 on their respective OS and duke it out.
Something not related to the pathetic troll Detox - one other thing to consider is that a different HDD was used for comparison (not different physical drives, but different manufacturers and models). I am not sure which is better, but they should have used the same model drives (GB size does not make drives equal) if you are going for an exact comparison. Also, I believe running under bootcamp requires OSX to be on the same drive, which could have caused fragmentation issues.
I actually didn't even read anything you wrote, NoHOne. It was just too long and talking to you yesterday I can see that 90 percent of what you say is BS anyway, so why should I bother?
It just seems you have an overreaction to anything that may put Windows or Microsoft in a bad light, especially if it's about Apple. You just seems so desperate, that's all.
Calm down. The world won't end tomorrow, even when it gets filled up with Apple products. ;)
@Detox - No, I have an adverse reaction to stupid comments and stupid people. If you present an explanation that I disagree with, then I will be happy to debate you and be respectful. But if you do the fan boy thing, attack without provacation (such as refusing to read a 16 lines of text but attacking), throw out things that are blatantly false when it is obvious you did not try to learn the truth, etc. (you did not do all these things, they are in general) then I will call you out as the fanboy you are. And as long as I don't overstep the rules of this forum, I will happily do it.
A proof of this is what Tseng wrote. Usually he does the usual bash thing without being provoked. But here, there was nothing like that, he made a simple statement I disagreed with so I didn't attack him - in fact, what I wrote was giving an idea as to how the numbers could benefit Apple's numbers.
But here you are, trying to play the "Apple and their supporters are being attacked" card, and yet I look through your posts in this article, and all I see is you calling people haters, MS apologists, dicks, and so on. But us MS supporters are supposed to shut up, and allow you to say whatever you wish.
For those looking for a device strictly for reading, the new Kobo is a nice little option. It's small enough to slip into a pocket, can do more with a PDF than the competition, and at $129, it's $10 cheaper than both the Nook and Kindle WiFi.
The most commented posts on Engadget over the past 24 hours.
Now that we've thrown 'em off the trail, use the form below to get in touch with the people at Engadget. Please fill in all of the required fields because they're required.
This could get ugly pretty quickly.
I would like to point out that some of the tests on both systems are done with Quicktime and iTunes, which are really testing how well Apple ported the applications to Windows than the actual speed (and the Windows ports are definitely worse in terms of performance).
Indeed. This benchmark is bullshit. At most, it highlights how poor Apple's porting skills are.
Do some tests with some real software. Encode some MP3s with LAME on either OS and see how it fares.
So 1/4 benchmarks were done with Apple software and that makes it a biased test? What are you smoking and where can I get some?
@Mio: EXACTLY! When I recently benchmarked my Core i7 Hackintosh against some other computers, I looked no farther than FFMPEG. Multi-threaded, multi-platform -- I could easily compare Apples to Oranges, erm, Windows to Ubuntu.
Actually, there are a number of items that could sway the numbers
First, Win7 64bit is a pure 64bit OS. SL requires you to hold down the 6 and 4 keys to boot as 64bit. Was this done? The article does not say.
For Windows to run, Apple has provided drivers. Are these Apple written drivers optimized for Win7 64bit or are they the older Vista drivers? Drivers that are a few years old compared to drivers that had optimizations performed over the years would have an impact. I could say that Apple may have written their Win drivers to perform slower, but I don't think (or I hope) they are not that evil. This would impact startup performance since those drivers need to load during startup, and impact shutdown since the drivers need to be closed, and have a performance hit over the time use of the computer.
They used QTX for SL, which the article states "Apple claims to have much improved performance over the previous version." Since QTX is not available for Windows, the numbers are going to go in favor of OSX. If you are using an optimized newer version vs. an older slower performing version, then you are going to have the results skewed towards OSX. They should have used software such as Zune 4.0, which not only is the newest version, but also is a 64bit app throughout. Since QTX is 64bit onSL and QT for Win7 is 32bit, you are going to have different results as encoding/decoding can take advantage of 64 bits, but can't when using QT on Win7.
Of course, this are real-world use so you need to take the numbers as they are if you were using a Mac as your hardware, but it would be interesting if you compared neutral software, or compared WMP/Zune vs. iTunes for those numbers (WMP would be less desirable as it is a 32 bit app).
There isn't a way to make this comparison valid, really. Apple builds the OS and the machine to work well together. Microsoft builds the OS to work on a huge variety of machine configurations. Since Microsoft doesn't have retail machines that Windows 7 was specifically designed for, the comparison falls apart, even if you put both OS'es on a third-party box.
You dont have to hold down the 6 and 4 to run everything just as fast and in 64 bit mode.
@NoHone
I agree with you for the most part, but wouldn't multi-platform apps such as ffmpeg or vlc be more of fair comparison? There is a lot more to learn from fair benchmarks, like how the Macbook Pro got the highest Vista system rating in 2008.
I might add that starting a Mac in 64 bit mode only affects the kernal, the OS and apps still run in 64 bit even when kernal is in 32 bit mode.
@BillG
What fair test would that be?
@Mio
Yeah complete bullshit, for consistency lets look at this:
Different HDs even if they are spin at the same rate, have the same interface, and same amount of cache they are not the same HD.
Brand new version of iTunes for SL and version 7.6.4 for Windows? Where is version 9 on Windows?
I have a more balanced comparison, how about you run your OSX/Win7 tests on the MBP and then you go get a Win7 based machine that costs the same amount as the MBP and compare that machines performance. Don't forget the pretty graphs.
@CtrlBurn - This is not exactly true. While they can pick and choose which hardware they use, for the most part all the hardware in a Mac is the same as a Windows computer. Really it comes down to the drivers in how the OS interacts with the hardware. The only place where the hardware has a real effect is that Mac uses EFI while Windows uses an older, slower BIOS. This will benefit SL and negatively impact Win7 as (if I remember correctly) BIOS is emulated when using Mac hardware.
@Tsing and BillG- Actually, if they boot as 64 bit it would benefit Mac as the kernel used is 32 bit unless you do the 6 and 4 key thing. Using a 32 bit kernel, which uses 32 bit drivers can, under some circumstances, adversely be a performance bottleneck compared to a 64 bit kernel with 64 bit drivers. For example, when they are reading from the CD/DVD drive, if they can read 64 bits at a time rather than reading 32 bits it would help perf. If the encoder needs to go to a driver or the kernel, there is a 64 to 32 bit thunk (using Windows terminology) and back which can be a perf bottleneck.
@BillG - which is why I wrote "it would be interesting if you compared neutral software"
Something else I just thought of that could skew the results (again, this it is fair since it is real world use, but an explanation of results), shouldn't the results be even better for iTunes on SL being they touted they supported CUDA (or whatever they were calling it), where encoding/decoding would give a great benefit? Or do they not use it in QTX?
Good lord, that NoHOne dude is so insecure about these results that he decided to write us a mini-novella to tell us about it. I would hate to see what he would write once they release the benchmarks for Macs with LightPeak.
Calm down, man. It's not the end of the world if Snow Leopard is a little faster. You'll live.
Amen to that. iTunes and Quicktime run beautifully on my Mac. They run like total ass on Win7. Every other app under the sun runs beautifully on Win7.
Sometimes I wonder if Apple intentionally makes their Windows applications bad to try to make a really childish point.
@Detox - you really are a sad person, aren't you? I am starting to believe, as mentioned yesterday, thay you are Paul being Paul wrote that same exact thing to me once, word for word.
If you were not such a pathetic Apple fanboy, you would see where I am saying things that would help SL have better numbers to bring them up to a more equal footing. You would also see where I am saying that there is no harm in using the most optimized software for each platform. If you want an exact performance comparison, then yes, you should use the exact same software with the only difference being the underlaying OS. But if you want to use the OS that gives you the best overall perf, then use iTunes with QTX vs. Zune 4.0 on their respective OS and duke it out.
Something not related to the pathetic troll Detox - one other thing to consider is that a different HDD was used for comparison (not different physical drives, but different manufacturers and models). I am not sure which is better, but they should have used the same model drives (GB size does not make drives equal) if you are going for an exact comparison. Also, I believe running under bootcamp requires OSX to be on the same drive, which could have caused fragmentation issues.
I actually didn't even read anything you wrote, NoHOne. It was just too long and talking to you yesterday I can see that 90 percent of what you say is BS anyway, so why should I bother?
It just seems you have an overreaction to anything that may put Windows or Microsoft in a bad light, especially if it's about Apple. You just seems so desperate, that's all.
Calm down. The world won't end tomorrow, even when it gets filled up with Apple products. ;)
.......
@Detox - No, I have an adverse reaction to stupid comments and stupid people. If you present an explanation that I disagree with, then I will be happy to debate you and be respectful. But if you do the fan boy thing, attack without provacation (such as refusing to read a 16 lines of text but attacking), throw out things that are blatantly false when it is obvious you did not try to learn the truth, etc. (you did not do all these things, they are in general) then I will call you out as the fanboy you are. And as long as I don't overstep the rules of this forum, I will happily do it.
A proof of this is what Tseng wrote. Usually he does the usual bash thing without being provoked. But here, there was nothing like that, he made a simple statement I disagreed with so I didn't attack him - in fact, what I wrote was giving an idea as to how the numbers could benefit Apple's numbers.
But here you are, trying to play the "Apple and their supporters are being attacked" card, and yet I look through your posts in this article, and all I see is you calling people haters, MS apologists, dicks, and so on. But us MS supporters are supposed to shut up, and allow you to say whatever you wish.
Snow Leopard does NOT require 6 and 4 to be held down during boot to start in 64-bit mode. That only forces 64-bit mode for lesser-performance Macs.