MacBook Benchmarking in Windows XP

In the computer world, desktops are nearly always faster than laptops. However with the onset of Intel’s Core Duo processor with the 945GM chipset, for the first time in my life I can say my laptop is faster than my desktop. I was very curious as to how this new MacBook would perform when put to the test and I was very surprised with the results. While I knew right off the bat that the GMA 950 graphics chip on the MacBook would hold it back, I had faith that the Core Duo processor would be more than a match for my ancient Athlon64 3200+.

When I was going over my list of benchmarks I noticed a severe lack of CPU benchmarks. I originally planned on just running a few iterations of Super Pi and then running through the 3DMark Series and calling it a day but I knew this wouldn’t show the true potential of the MacBook. After looking around a bit I found the PCMark series of benchmarks which fit the bill just perfectly. Unlike the 3DMark Series, PCMark targets the entire computer and stresses every component of the machine, not just the graphics subsystem. Now that I had my benchmarks, it was time to prepare the machines and watch them grind through the tests.

Test Configuration

Since the MacBook would be operating at stock frequencies I figured it would only be fair to run the desktop at stock frequencies. My MacBook has a 2GHz Core Duo, 2×256MB of ram, an 80GB 5400RPM hard drive, and an Intel GMA950 graphics chip. My desktop has a 2GHz Athlon64, 2×512MB of ram, 250GB 7200RPM main drive, and an ATi x700 Pro with 256MB of video ram. While the hardware isn’t fair, what the MacBook lacks in graphics it makes up in processor power. Both systems are running nearly fresh installations of Windows XP SP2 with as many processes killed as possible. Once I had installed all the benchmarks on both machines, I defragmented the hard drives, restarted the computers, defragmented again, restarted one final time, and then killed all extraneous processes. Once both machines had reached an idle setting it was time to start the tests.

Super Pi 1.4

I consider Super Pi to be the most accurate program in determining a computer’s sheer processing power. Super Pi is a very simple program that uses a processor to estimate Pi to a specific decimal place. Because you are estimating a floating point number and performing many iterations of the same computation you are heavily stressing the Arithmetic Logic Units that reside on the core. With Super Pi, the only important factor is raw computing power, memory bandwidth plays a small factor while graphical and storage abilities do not matter at all. For this test I calculated Pi to 1, 2, 8, and 32 million decimal places. For each calculation I would calculate Pi then wait a minute and calculate Pi again. In every case, the second calculation was a few fractions of a percentage point faster and I feel that the fastest times should be shown.

If these numbers do not mean anything to you just take notice by how much faster the MacBook computes when compared to the desktop. To stress how much faster the MacBook is than my desktop, I had to overclock my desktop to 2.8GHz just to match the MacBook performance in Super Pi 1M. This performance is all available on a platform that doesn’t emit much heat and has a great battery life.

3DMark Series

3DMark is the premier benchmarking system for testing your graphical subsystem. This benchmark has been around for the last 6 years or so and has always provided a great benchmark for the gamer and tweaker. Every new revision to 3DMark incorporates new technology that will be seen in the next generation of games. While it may make no sense to test current cards with future technology, by doing this you make sure that the cards are pushed as far as possible which current software may not be capable of. With 3DMark I already know that the desktop will come out with a large lead but I am performing this test just to show where the MacBook stands in regards to gaming.

The numbers just don’t do this benchmark justice. When performing it on my desktop, all but 3DMark 06 looks gorgeous. However the same thing can’t be said for the MacBook. That GMA950 was struggling on every benchmark except 3DMark 01. While these benchmarks may have looked terrible on the MacBook, it is still capable of doing some moderate gaming. I have been told that it runs CounterStrike:Source with decent graphics and textures and it runs World of Warcraft like a charm. Just don’t expect to crank up the settings on video games and expect it to run smooth as glass. One note, these results can almost be taken as results for a 2GHz Mac Mini since the graphics chip is identical and the chipset is the same.

PCMark Series

Unlike 3DMark, PCMark focuses on stressing the entire computer by using what most people would consider to be normal applications. PCMark goes about encrypting files, compressing and decompressing files, encoding audio files, rendering webpages, and accessing the hard drive. I feel that of the three benchmarks, this benchmark should be taken most seriously since it directly tests what most people will use the MacBook for. One note, I tried to get PCMark05 to run on the MacBook but it would only finish on my desktop, either the Windows Media Player installation screwed up or the Windows Media Encoder didn’t work correctly.

These are the results I wish I had seen with 3DMark. I am pretty certain that if PCMark05 had completed on the MacBook then there would be even more results showing the true power of the MacBook. The dual core processor coupled with higher memory bandwidth really help the MacBook to crunch the numbers.

Conclusion

I think these benchmarks really show the full spectrum of what the MacBook is capable of. While it doesn’t excel at gaming or drawing pretty polygons, it has more than enough power to last a few years with OS X and Windows. If everything works well in Vista I have a feeling this laptop will be good to go till maybe 2008 or 2009 unless there is a major shift in operating system requirements. Now my only gripe with the laptop would be the GMA950 but I knew what I was getting before I purchased the system. If you are an avid gamer and are looking to make this laptop an upgrade I would suggest checking out the MBP. However if you just need a light laptop with a great battery life, look no further than the MacBook.

The Conversation {5 comments}

  1. scottfrye {Monday June 12, 2006 @ 9:23 am}

    Do you know about how much the MacBook weighs?

  2. Chris Morrell {Monday June 12, 2006 @ 9:41 am}

    Apple rates the MacBook at 5.2 pounds and the MacBook Pro at 5.6 pounds for the 15″ and a hefty 6.8 pounds for the 17″. As for battery life the MacBook is rated for up to 6 hours while the 15″ MBP is rated for 4.5 hours and the 17″MBP is rated for 5.5 hours. So far with my Apple experience I can say the figures for the MB seem to be right on the dot with everything working perfectly.

  3. L A {Sunday December 24, 2006 @ 2:51 am}

    Macbook Pro 15″
    2.33 Ghz core 2 duo
    256MB ATI mobility x1600
    5400RPM 160GB hitachi

    Pcmark05 –> 4443
    3dmark05 –> 4150
    3dmark06 –> 2010

  4. Igor {Thursday January 25, 2007 @ 1:43 pm}

    Hi there!

    Great review, however, one thing that I really miss on all reviews with 3dmark is that only the final score is show. It would be great to have the individual scores, for CPU and GPU, because that is what MBP lacks, 3d power, but processing power, well, that it has. =)

    Thanks for the nice article.

    []’s

  5. Chris Morrell {Thursday January 25, 2007 @ 1:49 pm}

    That’s a very good point Igor, when I did the testing I should have saved the various results as it would have shown exactly how much performance the GMA950 was adding to the overall score. If I can snag one of the new C2D MacBooks then I will surely compare the two and post the full results.

Speak Your Mind

  • Comment Policy: No flaming unless directed at the author for good reason. All other flaming will be turned into a comment about ponies and rollerskates.