SSD Performance on Macbook Pro

Posted by Brenden Cyze Fri, 10 Dec 2010 08:58:00 GMT

I have an older model Macbook Pro (Model 3,1 2.6 GHz 17”) that I wanted to refresh before I considered investing in a brand new one.  I successfully increased my RAM to 6GB with a noticeable effect.  However, software builds were still taking quite awhile, so I replaced my existing 7200 RPM hard drive with an SSD.  The post is an unscientific summary of my experience.  Before purchasing a SSD, you need to determine if your specific computer usage pattern coincides with one or more of the many SSD device disadvantages.


The following table summarizes my Xbench performance metrics.  This is a representative sample of the performance numbers based on a dozen benchmark snapshots.  The table nomenclature is HD==hard drive; SSD==solid-state drive ; Change==Percent Change = (SSD - HD)/HD x 100; the Disk Test is described in the Xbench documentation.

Xbench disktest hd vs ssd











I was amazed at the performance boost for “random” disk operations (reads/writes).  I highlighted the last two lines since they are represent some of the SSD hype drivers.

You might wonder why I have quietly ignored the 6,528% increase in uncached writes.  Since the SSD write performance degrades overtime, I wanted to table any analysis of this metric until I have had more time to churn the contents of my SSD.

The biggest challenge with interpreting any benchmark framework is that they amounts to a lab experiment.  Most people use their notebook (desktop) in a mixed workload manor.  Of course, there are categories of users:  developers, power users, etc.  However, the majority of interactions are not pure read or write operations nor absolutely random.

I think that this is one of the biggest reason why the overall “performance experience” may not meet (should exceed) expectations.  That is, overall performance is typically muted from all the noise that the benchmark framework skillfully factored out.

I recall the sustained performance boost (gave me an additional 14 months of useful life before I handed it down to a family member) I received when I upgraded my previous notebook with a 7200 RPM drive.  In this case, the experience has been positive; however,  I cant say that I received the utility I expected from such an expensive investment.  Also note that the MBP 3,1 model’s controller throughput is 50% that of the newest MBP model - I have absolutely no doubt that if I had dropped this device into a brand new Mac Book Pro the performance would have been amazing.

Hardware Specifications

Posted in  | Tags  | no comments


(leave url/email »)

   Comment Markup Help Preview comment