Geek Patrol

MacBook Pro Running XP Geekbenched

I noticed this morning that we’ve started receiving Geekbench Preview 2 results for MacBook Pros running both Mac OS X and Microsoft Windows XP. I thought it would be interesting to compare the performance of the same hardware (in this case a MacBook Pro) running different operating systems.

Setup

Here is the configuration of the test machine:

• MacBook Pro
  • Intel Core Duo 2.0GHz
  • 1GB RAM
  • Mac OS X 10.4.5 (Build 8H1455) or
  • Windows XP Professional

We’re using the baseline scores (not the raw scores) from each benchmark, so in all cases a higher score is better. We’ve computed the Windows XP score as a percentage of the Mac OS X score, and as always, higher is better.

Results

Conclusion

Most of the benchmarks in Geekbench are strictly CPU and memory benchmarks that don’t really rely on the operating system (the bzip2 and stdlib benchmarks being the only exceptions); the performance of these “simple” benchmarks depends more on the code the compiler generates rather than the operating system the code is run on.

This leads me to believe that the results of the Blowfish benchmarks (where the Windows XP MacBook Pro was 10 times slower than the Mac OS X MacBook Pro) are a result of the compilers used (in this case Visual C++ and GCC) rather than the operating systems used.

Of course, Geekbench also has benchmarks that do rely on the operating system; the stdlib benchmarks, in particular, derive almost all of their performance from the standard library implementation of the operating system. Here we see two benchmarks (stdlib allocate and stdlib copy) where Windows XP is faster than Mac OS X (at one point, Windows XP is 35 times faster than Mac OS X), and another (stdlib write) where Mac OS X is faster than Windows XP.

Overall, there are areas where the Windows XP MacBook Pro was faster, areas where the Mac OS X MacBook Pro was faster, and areas where they were both roughly the same. Looking at these results, it’s hard to say which configuration comes out on top, although I think you could make a convincing argument for Windows XP (with Visual C++) being a bit faster overall than Mac OS X (with GCC).

Nature Cooked Wikipedia Study

Remember the Wikipedia and Britannica comparison Nature published a few months ago? According to The Register, Nature might have cooked the study to favour Wikipedia:

But Nature sent only misleading fragments of some Britannica articles to the reviewers, sent extracts of the children’s version and Britannica’s “book of the year” to others, and in one case, simply stitched together bits from different articles and inserted its own material, passing it off as a single Britannica entry.

Nice “Mash-Up” – but bad science.

Bad News Day

Yesterday was not a good news day, at least for Apple and Microsoft.

In the case of Apple, the French Parliament has passed a bill that if also passed by the senate, will force Apple to either license FairPlay or pull out of the country all together. Apple is apparently (and understandably) not too happy about what the bill could mean.

While I understand the implications of this I don’t really have an opinion. I still buy music on those old fashioned shiny plastic disks. They’re called “see-dees,” (you may have heard of them) and they have their own DRM issues.

In the case of Microsoft, that shiny new operating system we keep hearing about has been delayed until 2007. The version for business will still come out in November 2006, but we consumers shall have to wait until the following January.

While reactions have been mixed, I can’t say I am terribly surprised. I mean, it might be a stereotype, but Microsoft has made a habit of pushing things back. The question is, will consumers buy the “security enhancements” reason that Microsoft has given.

Geekbench FAQ

We’ve created a Frequently Asked Questions page for Geekbench. Please feel free to peruse it. If you have further questions, please feel free to ask them. Check out the about page to find out how to get in touch with us.

We’ve Moved

I thought I’d let both of you know that we’ve switched hosting companies. If you’re reading this then that means you’re visiting us on our new server. Hopefully everything made it in the transition and everything “just works”, but if it doesn’t, let us know.

Windows on Mac Hardware, No Mac on Generic Hardware

For those of you who are interested, someone has managed to get Windows XP to boot on an Intel Mac.

The first hint that “narf” may have figured it out came in the form of a set of photos on Flickr that seemed to show Windows XP on an iMac of some sort. Next came the videos and an acknowledgement from the contest administrators that a solution had been submitted. And finally, this morning, onmac.net announced that a proposed solution had been found and had been tested successfully. The contest had been won, and it looks like “narf” is about to almost US$14,000 richer.

The WindowsXP on Mac website has now posted the solution developed by Narf. It is fairly complicated and will probably be a little difficult for novice users to get a handle on, but I’m certain that an easier, more accessible way to deploy this will become available in a relatively short amount of time. The only downfall that I can see is that you’ll need a PC to complete the process, which defeats the purpose of this exercise for a portion of the people who’ve been anticipating this solution.

And, as a counterpoint, here is an interesting article about why OS X shouldn’t be sold for generic PCs

The appeal of the Mac OS is successful integration of it with the hardware; something that is available only when one company owns the whole setup.

Now, realistically, could this be accomplished for the PC world? Well sure. Just buy one, each, of the following:

1. Every motherboard sold since 2000.
2. Every sound card sold since 2000.
3. Every video card sold since 2000.
4. Every processor sold since 2000.
5. Every CD-ROM sold since 2000.
6. Every keyboard, mouse, joystick, and touchpad sold since 2000.

Then one of each of the following:

1. Every complete desktop system sold since 2000 by a major manufacturer (Dell, Compaq, Gateway, HP, etc.).
2. Their portables, too.

Now, write support for them all in the OS to the level the Mac has support of all their computers back to 2000. At this point we can safely say that the Mac OS would be equal on PC to the Mac OS on the Macintosh. It will also occupy three DVDs and have piss-poor support for half of the devices.

Interesting times in the Mac world.

iMac Geekbench Comparison

12 different iMacs Geekbenched.

No Vista for You

For those of you wanting to install Windows on your shiny new Intel Macs, it looks like Vista will not be able to install and run on them. Well, at least not the iMac, Mac mini or x86 PowerBook[1].

According to C|Net:

One of the big obstacles is that although both the Macintosh OS and Windows now use Intel chips, the two operating systems have different ways of booting up.

Mac fans have held out considerable hope that the next version of Windows would be easier to load on Macs than Windows XP, because like Mac OS X, Vista will use Extensible Firmware Interface (EFI) to aid the boot process. Older versions of Windows have used Basic Input Output System (BIOS).

However, Apple Senior Software Architect Cameron Esfahani said that his understanding is that only 64-bit versions of Vista will support EFI. To this point, all of the Intel Macs have used 32-bit chips.

“I don’t think so” he said, when asked whether Intel Macs should easily run Vista. Esfahani was speaking at the Intel Developer Forum here–to a packed room despite the session being the last of the conference.

In addition, while EFI has prospects for supporting many older types of software and hardware, Apple has not included much of that “legacy” code in its EFI implementation.

“Windows is a legacy OS,” he said to laughter and applause from the crowd. “We don’t have legacy support.”

That’s not to say there isn’t hope, though. 64 bit editions of Vista will support EFI, and PowerMacs and Xserves will likely (in other words, had better be) running 64 bit Intel CPUs.

So it’s not that your dream of running Windows has been crushed, it has just been delayed.

[1] I still don’t like the name ‘MacBook Pro.’ Also? I’m stubborn.

Mac minis Geekbenched

We’ve started receiving Geekbench Preview 2 results for the new Mac mini Core Solo and Duo. Since Matt and I have an unhealthy obsession with benchmarks, we thought both of you might be interested in seeing how the new Mac mini compares to the old Mac mini.

Here is the configuration of our three test machines:

• Mac mini G4 (old and busted)

  • PowerPC G4 1.5GHz
  • 1GB RAM
  • Mac OS X 10.4.5 (Build 8H14)

• Mac mini Core Solo (new hotness)

  • Intel Core Solo 1.5GHz
  • 512MB RAM
  • Mac OS X 10.4.5 (Build 8H1619)

• Mac mini Core Duo (new hotness times two)

  • Intel Core Duo 1.67GHz
  • 512MB RAM
  • Mac OS X 10.4.5 (Build 8H1619)

We’re using the baseline scores (not the raw scores) from each benchmark, so in all cases a higher score is better. We’ve computed the Mac mini Core Solo and Duo scores as a percentage of the Mac mini G4’s score, and again, higher is better.

The only benchmark where the PowerPC-based Mac mini outperformed the Intel-based Mac mini was Stdlib Allocate, which is also the only benchmark where the PowerBook G4 outperformed the MacBook Pro. The Stdlib Allocate benchmark depends more on standard library performance than raw hardware performance.

It’s also worth mentioning that the Intel-based Mac minis were, for the most part, faster than our baseline system (a PowerMac G5 1.6GHz). Yikes.

Overall, we’re impressed with the Intel-based Mac minis’ performance. The only area we’re concerned with is graphics performance (due the the Intel-based Mac minis using an integrated graphics adapter) which, unfortunately, Geekbench doesn’t measure. Otherwise, the Mac mini is the bee’s knees.