Aperture, Apple, Tech

Aperture Benchmarks between Quad G5 and MacBook Pro

The MacBook Pro (MBP) is my first Intel Macintosh. For the past couple of years, I’ve been using a Quad G5 as my primary machine and a first-generation PowerBook G4 12-inch for travel. Before I purchased the Quad G5 in 2005, my desktop was a Quicksilver PowerMac G4, purchased in 2001. Though the PowerMac G4 ran at the same speed (867MHz) as the PowerBook (purchased in 2003), it was faster overall due to a better graphics card, more RAM, and a faster hard disk.

It’s been two years, and here I am with a new MacBook Pro. Unfortunately, it’s not the new Penryn-powered MBP; I was on a plane to Hong Kong just days before the new ones were released. Once there, there was no way to take advantage of Apple’s 14-day return policy.

With a little time on my hand, I wanted to see if the MacBook Pro’s performance was up to par with the last of the great PowerPC Macintoshes. In addition, I’ve been curious for some time about about Aperture’s performance on Intel and Motorola PowerPC chips.

I conducted three tests using Aperture 2.0.1 and a sample library containing twenty RAW photos from an 8 megapixel Canon EOS 1D Mark II. Ten of the images had no adjustments, keywords, or metadata applied to them. The other ten images had keywords and metadata, in addition to white balance, exposure, enhance, and edge sharpening adjustments.

My Quad G5 features 6GB RAM, an Nvdia GeForce 7800GTX with 512MB of RAM, and a 750GB Seagate internal SATA hard drive. The MacBook Pro has 4GB RAM, an Nvdia GeForce 8600M GT with 256MB of RAM and a 160GB Hitachi internal SATA hard drive.

All tests were performed four times, and the results are averaged in the tables below:

Test 1: Export Ten Non-Adjusted Images to 16-bit TIFF

For the first test, I selected the ten images that had no adjustments and exported them to 16-bit TIFF files. Normally, this isn’t something I would do, but I wanted to see how fast Aperture could export without doing much processing.

Computer Total Time (seconds) Time Per Photo
Quad G5 25.30 2.53
MacBook Pro 28.80 2.88

Not bad. The MacBook Pro more than held up against the mighty Quad-core PowerMac in this test with the Quad coming in at 12% faster over the MBP. Since I don’t often export the files without performing any adjustments, I am more interested in the results from the next test.

Test 2: Export Ten Adjusted Images to 16-bit TIFF

Each image had its white balance, exposure, enhance, and edge sharpening adjusted. This is a better test, because the computer’s CPU has to perform additional calculations before exporting the image. If I had another video card for the Quad G5, I could see whether or not the GPU made a difference. In my Quad, I am using a reflashed Nvidia 7800GTX with 512MB of RAM. The original stock 6600GT card was such a dog that it’s no longer in my possession.

Computer Total Time (seconds) Time Per Photo
Quad G5 91.38 9.14
MacBook Pro 117.10 11.71

The Quad G5 is 22% faster than the MacBook Pro. When exporting hundreds or even thousands of photos, every percentage point counts. For a typical wedding shoot, I might be exporting 1000 images. The Quad G5 would take two and a half hours, whereas the MBP would take an additional forty five minutes.

Test 3: Export 1024 x 1024 JPEG

Exporting a web-sized or preview sized JPEG is a common task for me. Here I exported the ten adjusted images to fit within a 1024 x 1024 pixel box.

Computer Total Time (seconds) Time Per Photo
Quad G5 87.95 8.80
MacBook Pro 122.28 12.23

I was a little surprised by this score. While the Quad was faster at exporting an 8-bit JPEG over a 16-bit TIFF file, the MacBook Pro was slower. Exporting 1000 images, the Quad would take 2:26:00 while the MBP would take nearly an hour longer at 3:24:00. I wonder if the Quad was created the JPEG from its pre-generated preview image.

So, after two and a half years, the Quad G5 can still hold its own against a MacBook Pro. The gap might be a tiny bit closer, however, with the new Penryn-powered MacBook Pros. I’m curious to see how much faster the Dual 3.2GHz Quad-Core “Harpertown” MacPro is compared to the Quad. The MacPro would have to be twice as fast as the Quad before I would even consider upgrading. When I went from the Quicksilver to the Quad G5, I saw a 600% speed increase in RAW processing performance. Yes, the Quicksilver took about a minute to process one RAW file!

Another test I’ve been conducting is power per watt. With a Kill-A-Watt power meter, the MacBook Pro is well ahead of the Quad G5, which is one power-hungry computing beast!

Standard

One thought on “Aperture Benchmarks between Quad G5 and MacBook Pro

  1. Joe says:

    Thank for this interesting comparison. I’m a bit surprised that the current Santa Rosa Macbook Pro was actually slower then the G5 when people around the net claim that the MBP was actually just as fast or even faster then the G5. Still it is quite impressive that a laptop could keep up with the once state of the art workstation.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s