Windows XP on Intel Mac benchmarks posted online
Benchmark tests of Windows XP Pro running on all three of the first Intel-based Macs reveal that the MacBook Pro runs Adobe Photoshop faster than other laptops originally designed for Microsoft Windows.
With the help of the recent Windows XP for Mac software hack, a team from Gearlog was able to install Windows XP on a 20-inch iMac, MacBook Pro, and Mac mini Intel Core Duo. It then ran benchmark tests to gauge the speed of the Microsoft operating system on Apple's hardware.
To spice it up a bit, they allowed all three machines to remotely access a fourth Mac system through a VNC client -- effectively displaying the Mac OS X interface through a window on Windows XP systems.
"The MacBook Pro is the fastest Core Duo laptop we've tested running the Photoshop scripts. It's faster than other laptops originally designed for Windows," the team said. "This bodes very well for the performance of an Intel-accelerated OS X Photoshop, when that finally appears."
Based on existing benchmarks, the MacBook Pro beat four other non-Apple Core Duo laptops on the Photoshop test, but came in behind them on the Windows Media test. Meanwhile, the two Mac desktops "outran even blazing-fast single core systems, which typically do the Windows Media Encoder test in 10-13 minutes."
According to the report, installation of Windows XP onto the Apple systems wasn't difficult, thanks largely to detailed guides available at the OnMac. However, each of the three systems required a different version of the xom.efi file -- the bootloader which lets the system choose between Windows XP and Mac OS X.
Due to a lack of video drivers for the iMac or MacBook Pro, the team did not benchmark video game performance, but they did get Ethernet, wireless networking, and the headphone jack (but not the internal speakers, iSight or the remote) working using drivers suggested by OnMac.
"Apple makes fast Windows PC," the report concluded.
With the help of the recent Windows XP for Mac software hack, a team from Gearlog was able to install Windows XP on a 20-inch iMac, MacBook Pro, and Mac mini Intel Core Duo. It then ran benchmark tests to gauge the speed of the Microsoft operating system on Apple's hardware.
To spice it up a bit, they allowed all three machines to remotely access a fourth Mac system through a VNC client -- effectively displaying the Mac OS X interface through a window on Windows XP systems.
"The MacBook Pro is the fastest Core Duo laptop we've tested running the Photoshop scripts. It's faster than other laptops originally designed for Windows," the team said. "This bodes very well for the performance of an Intel-accelerated OS X Photoshop, when that finally appears."
Based on existing benchmarks, the MacBook Pro beat four other non-Apple Core Duo laptops on the Photoshop test, but came in behind them on the Windows Media test. Meanwhile, the two Mac desktops "outran even blazing-fast single core systems, which typically do the Windows Media Encoder test in 10-13 minutes."
According to the report, installation of Windows XP onto the Apple systems wasn't difficult, thanks largely to detailed guides available at the OnMac. However, each of the three systems required a different version of the xom.efi file -- the bootloader which lets the system choose between Windows XP and Mac OS X.
Due to a lack of video drivers for the iMac or MacBook Pro, the team did not benchmark video game performance, but they did get Ethernet, wireless networking, and the headphone jack (but not the internal speakers, iSight or the remote) working using drivers suggested by OnMac.
"Apple makes fast Windows PC," the report concluded.
Comments
Consumers can have a system that runs OS X and Windows. I think it's a no-brainer.
Can't wait for PhotoShop to be Universal so we can compare the MacIntel version to the Windows version.
Originally posted by Xool
Can't wait for PhotoShop to be Universal so we can compare the MacIntel version to the Windows version.
Yeah, that's the only test of interest.
Comparing Windows apps running on a Mac to Windows apps running on a Dell makes no sense. If they have the same components, they're running the same OS and the same software, you don't need a benchmark to know that they're the same. If one has a faster chip/bus/memory/drive, then we know that machine is a faster machine by the amount that its components influence speed. What's the point? The Apple label isn't going to make the machine faster or slower.
Originally posted by Xool
Can't wait for PhotoShop to be Universal so we can compare the MacIntel version to the Windows version.
And of course to the PowerPC version also. So, which will win? The real test is going to be interesting as there's a switch to a completely new compiler as well as architecture.
Is CS3 going to be faster than CS2 on the G5?
Is the XCode compiled CS3 on Intel Macs going to be faster than on Windows?
Is the G5 going to be faster than Intel anyway because of AltiVec?
Worst case scenario for Apple is that CS3 is slower than Windows and slower than the G5 and that's not entirely unlikely either.
Worst case scenario for Apple is that CS3 is slower than Windows and slower than the G5 and that's not entirely unlikely either.
I don't see how this would be Apple's fault.
Its on Adobe to optimize CS for OS X.
A. Microsoft employees are excited about OS X.
From mini-Microsoft blog.
Yeah, you may have heard the argument I'm about to make before as it was used in reference to emulators, but emulators had a huge stumbling block that crippled the argument: performance.
So why *could* this be bad? Well, it could marginalize Mac software developement. Games is the chief sub-set of development that comes to mind. If all new Intel-Macs can easily dual-boot into Windows, and have appropriate drivers, then what incentive is there for Mac developers? People will not wait for a Mac-compatible version of Half-Life3 or DOOM-Everyone-Dies (TM) they'll buy the PC versions, dual-boot and play it to death, and once the Mac version is released, who is going to buy it? Further, for those game divisions that actually think of the Mac as they develope their Windows game (are there any these days?) what incentive is there for them to bother thinking about the Mac at all? After all, portability adds complexity, and if you know your customer is willing to dual-boot to play the latest-n-greatest, then why spend the development effort?
I'm not saying this is going to happen, and I'm not trying to be a wet-blanket here, but it's a genuine concern that I have. Emulators were never a threat to independant Mac-software development, because they were a mickey-mouse-duct-tape-and-wire last resort to using some application you *needed* while maintaining your Mac lifestyle. Full hardware compatible support however changes that.
So what do you think?
Just a thought...
Originally posted by fuzz_ball
Games is the chief sub-set of development that comes to mind. If all new Intel-Macs can easily dual-boot into Windows, and have appropriate drivers, then what incentive is there for Mac developers? People will not wait for a Mac-compatible version of Half-Life3 or DOOM-Everyone-Dies (TM) they'll buy the PC versions, dual-boot and play it to death, and once the Mac version is released, who is going to buy it? Further, for those game divisions that actually think of the Mac as they develope their Windows game (are there any these days?) what incentive is there for them to bother thinking about the Mac at all? After all, portability adds complexity, and if you know your customer is willing to dual-boot to play the latest-n-greatest, then why spend the development effort?
I understand what you're saying, but it seems to me that games are the perfect case for why this is a good thing. Right now, some people (OK, mostly kiddies, but they get a lot of computers) don't get Macs because they don't run the latest and greatest games. That would cease to be an issue. Furthermore, games don't really use the OS. You don't need Aqua for Doom. Doom looks essentially the same on Mac and Windows.
In the meantime, Apple has started creating a lot of well-respected software in addition to their well-liked OS: iLife and the pro versions of those apps in Final Cut Studio in particular. If people want to run those apps, they need to get a Mac. And if they have Windows apps (even games) that they don't want to lose, there's no longer a problem there.
Originally posted by fuzz_ball
So what do you think?
id makes their games available for Windows, Linux, and OS X, now that they are on the same architecture, I can only imagine that would become easier (two of them primarily running the same compiler doesn't hurt either).
Blizzard has long been faithful to the Mac community, and again, this shouldn't change anything (but as long as they still do OpenGL version of their games, why don't they just release a Linux version already???!!!).
I would argue that platform independence reduces complexity, because all the "complex" stuff gets abstracted out meaning the main program logic doesn't have to worry about it any more. As a programmer, I can say from experience, that program code I have worked on has gotten cleaner as the project moved from a single platform to multiplatform and environment support (not just other operating systems, but other support software as well). Development became easier, because we didn't have to worry about the details, they were handled in the platform specific part (which only needed to be written once per platform). That initial separation can be hard, but most big game companies are fairly mature.
The more Macs on the block, the more software vendors will be writing software for Macs.
Originally posted by fuzz_ball
So why *could* this be bad? Well, it could marginalize Mac software developement. Games is the chief sub-set of development that comes to mind. If all new Intel-Macs can easily dual-boot into Windows, and have appropriate drivers, then what incentive is there for Mac developers?
So what do you think?
I'm a bit concerned about Adobe-type apps, especially if Gates were to pull a fast one like offering Windows for $1 to Adobe to ship it with PhotoShop so that Adobe had no incentive to make the Intel-Mac version. But Jobs seems to have negotiated that away, and it would be bad PR for Microsoft.
Games - I have given up on them a long time ago. For 20 years we have seen the game developers make the Windows version and then months later, or not at all, grudgingly cough up the Mac version.
- The casual game-player doesn't have enough money to interest Apple anyway. And if they do, there might be just enough Mac games, if the game developers can figure out that parallel development, instead of the wasted energy involved in developing-then-porting, is better.
- The "gamers" do have enough money, but they build their own systems from parts from Newegg. They are not Apple customers anyway.
As long as the game developers continue to complain that 35 million users isn't a big enough market, nothing will change. Sure, their Mac sales are lousy, but what do they expect when they release the Mac version 6 months later and it is basically a hacked-up Windows port?
Originally posted by hohlecow
I would argue that platform independence reduces complexity, because all the "complex" stuff gets abstracted out meaning the main program logic doesn't have to worry about it any more. As a programmer, I can say from expuaerience, that program code I have worked on has gotten cleaner as the project moved from a single platform to multiplatform and environment support (not just other operating systems, but other support software as well). Development became easier, because we didn't have to worry about the details, they were handled in the platform specific part (which only needed to be written once per platform). That initial separation can be hard, but most big game companies are fairly mature.
The more Macs on the block, the more software vendors will be writing software for Macs.
Hear hear!
Breaking down the traditional barrier between the Mac and Windows worlds is great news for Mac sales and OS X. Developing for the Mac with Xcode and Cocoa is enough of a bonus all by itself when a company has a real choice between the platforms. And if Apple can make it easier to code multiplatform apps from square one with some real voodoo in the dev tools, they'd be making many people's lives easier other than just the Windows iTunes team!
As for the post about Microsoft doing a Wine ... it's not exactly in their interest. VPC on Intel is fine for them because it's still inconvenient compared to running Windows outright and sells another copy of the OS. A really integrated solution like Rosetta or Classic is something which only Apple and the F/OSS community can truly want. Don't expect wonders from Microsoft which don't extend their interests.
I don't think Microsoft are in a hurry to return to the days when they were a developer outfit dependent on Apple's machines!
Originally posted by TenoBell
I don't see how this would be Apple's fault.
Its on Adobe to optimize CS for OS X.
You're presuming that the compilers on OSX and Windows are equal and that the OS overheads are identical, which may or may not be the case. Regardless, it'll be embarrassing for Apple if otherwise identical applications run faster on windows on identical hardware or if the G5 beats the best Intel can offer.
Originally posted by JoeAlamaiz
I wonder, maybe Microsoft's next version of VirtualPC would be something like RedBox, if people are able to use Windows anyway via Double Boot, Microsoft might just allow people to do this through the OSX (in another layer, like "classic" or "rosetta"), if this means that you need to buy VPC + Windows... this came to my mind for some reason... I remember that someone at Microsoft said something about exploring the ways VPC should work in a MacTel... I think it was some lady at an apple keynote, hahah
It was Roz Ho, head of MacBU at Microsoft.
Dual booting isn't a real concern as it's pretty technical to set up and you've got to repartition a hard drive if you've not got a 2nd.
VirtualPC at native speed is the real concern because then you can run windows apps alongside your OSX apps without having to reboot. That means you can run AutoCAD or DOOM and the developer has zero incentive to port it to OSX. The second problem is that with seamless Windows app integration like that, alternative Mac applications aren't required. So, Windows developers have less incentive to port to OSX and OSX developers now have more competition and in particular it's from big bucks Windows developers. How is that a good thing for consumers or developers?
And by the looks of it the OSX port of QEMU will have native speed emulation well ahead of VirtualPC.
Originally posted by BRussell
I understand what you're saying, but it seems to me that games are the perfect case for why this is a good thing. Right now, some people (OK, mostly kiddies, but they get a lot of computers) don't get Macs because they don't run the latest and greatest games. That would cease to be an issue. Furthermore, games don't really use the OS. You don't need Aqua for Doom. Doom looks essentially the same on Mac and Windows.
It's probably good for Apple although most gamers wouldn't buy a Mac anyway, they'll build something out of bits. PC gaming has to be a dying market though with next gen consoles. At some point game developers are just going to quit the desktop game market.
Originally posted by BRussell
In the meantime, Apple has started creating a lot of well-respected software in addition to their well-liked OS: iLife and the pro versions of those apps in Final Cut Studio in particular. If people want to run those apps, they need to get a Mac. And if they have Windows apps (even games) that they don't want to lose, there's no longer a problem there.
The problem there is that those users will most likely continue to use Windows applications instead of replacing them with Mac alternatives.
Originally posted by AppleInsider
Due to a lack of video drivers for the iMac or MacBook Pro........
This was cuz of different firmware in the video chipset or something right? Man, I hope this doesn't mean that when the desktops start coming out, we'll still have to buy "Mac Edition" versions of the video card. I was hoping with the move to Intel, we'd be able to buy from the wide range of video cards out there for the generic x86 systems.
You're presuming that the compilers on OSX and Windows are equal and that the OS overheads are identical, which may or may not be the case. Regardless, it'll be embarrassing for Apple if otherwise identical applications run faster on windows on identical hardware or if the G5 beats the best Intel can offer.
No I'm not. I'm just saying there are more factors than the compilers or even the OS itself.
As far as we've seen Universal Applications have equaled or outperformed PPC native applications.
Developers have Intel Macs to test their Universal Apps, if Apple's Universal compilers were significantly inferior and offered significantly less performance than PPC native or Windows, the word would surely be out by this point.
Originally posted by TenoBell
Q. What's the difference between OS X and Vista?
A. Microsoft employees are excited about OS X.
From mini-Microsoft blog.
Originally posted by aegisdesign
You're presuming that the compilers on OSX and Windows are equal and that the OS overheads are identical, which may or may not be the case. Regardless, it'll be embarrassing for Apple if otherwise identical applications run faster on windows on identical hardware or if the G5 beats the best Intel can offer.
Compilers obviously won't be identical. Adobe is going XCode and the underlying GCC compiler is certainly not the fastest in the world (neither for compile- nor for runtime). But now at least Apple are on a chip platform where other companies such as RedHat are also contributing to GCC optimization. So improvements are ongoing.
For PowerPC most GCC improvements were done by Apple (and Linux versions for PowerPC could profit from that).