Number Crunching 10 years of Geekbench Scores

The new Sandy Bridge equipped Macs have all outperformed their previous generation counterparts in benchmarks by a significant margin. I thought it would be interesting to take a look at some specific product lines to see how they’ve performed over the years. To that end, I put together some charts comparing Geekbench scores for the iMac, the MacBook Pro, and the MacBook Air. For each successive model iteration, I used the highest available processor configuration1 to find the appropriate 32 bit Geekbench score. While useful for the purpose of this excercise, Geekbench scores presented here, like any specific benchmark, shouldn’t be considered an end-all, absolute reflection of real world performance.

Click through the links to view the charts.

Here’s the data for the iMac.

The MacBook Pro and the PowerBook G4.

And the MacBook Air.

(Some of the data was missing from the MacBook Pro Chart. The average percentage increase for the January 2006 release was 188%. The average percentage increase for February 2011 iMac update was 74%.)

So what can we glean from this data?

  • Not taking into account milestone processor upgrades(Highlighted in Blue), the average improvement for each iMac and MacBook Pro refreshes in the last 10 years are 14% and 12 %, respectively. 
  • Milestone improvements in processor lines take much longer on the MacBook Pro than the iMac. The notebook took 5 years to convert from PowerPC to Intel, and almost another 5 years to make the leap to a quad-core Sandy Bridge processor. Comparatively, the iMac made 4 generation leaps in the same time period, going through the G3, G4, G5, Core Duo and i7 lines. Perhaps more useful to us, it only took 3 years for the iMac to go from the Core 2 line to the i7.
  • Speaking of PowerPC, notice how the PowerBook lagged behind the iMac 300 points right before the introduction of Intel Processors. The difference would be greater if you compared it to its sibling pro level desktop. The Quad-Core 2.5 Ghz G5 PowerMac scored a 3318, nearly 300% higher. This goes to show how handicapped the PowerBook using PowerPC CPUs. This explains, even more than the inability to reach 3.0 Ghz, the need to switch to Intel. 
  • And what a difference the Intel Processors made. Performance per watts indeed.
  • I’d posit that scores within a processor family (PowerPC, Core 2, i7) tend to double every 3 years, give or take. The current crop of i5-i7 Macs analyzed, other than the Air, would find themselves in the middle of this 3 year cycle. The Air would be at the end of a cycle yet it has managed to nearly double its score in under one year. Might be useful to know if you’re considering an upgrade.

Some of this may not be news to you. If you’ve followed Apple for any reasonable amount of time, you’ll already be familiar with a lot of the history these charts characterize. For those of you who, like me, are newer to the Mac platform, maybe visualizing 10 years of Mac Geekbench scores can put into perspective how far we’ve come. 

As with the Mac Pro, comparing Geekbench scores for specific product lines can help choosing which machine will get us the most longevity for our money. While we all know a current iMac will offer more power than a current MacBook Pro, those who like to obsess over such things might semi-accurately compare the rate of performance improvements between the two. More information can never hurt.

I included the MacBook Air scores to help underline how early we are into the future of the next generation of MacBooks. Using a relatively nascent line of Intel Ultra Low Voltage processors (ULV), the Air is truly a product still in its infancy. Imagine where we might find ourselves in 10 more years. 


1. High end models tend to introduce new processors whereas mid to low-end models are often just the previous generations CPUs at a lower price. The goal is to try to see the difference new processors, although I suspect the results would be similar if you compared lower speced’ models to each other.