MacBook Pro Solid State Drive Upgrade Guide and Performance Testing

Performance Test Platform

The test platform is a base model MacBook Pro 13" from Early 2011 (Intel Core i5 2.3GHz, 4GB DDR3-1333).

  • Hitachi 5K500 320GB hard drive (5400rpm)
  • Western Digital Scorpio Black 500GB hard drive (7200rpm)

Solid State Drive Firmware Versions

  • Crucial m4: FW009
  • Kingston HyperX: 332
  • OCZ Vertex 3: 2.15
  • Other World Computing Mercury Electra 6G: 332

Testing Configuration and Software

  • InTech QuickBench 4.0
  • BlackMagic Disk Speed Test 2.1

A base OS X Lion 10.7.2 install is used for the operating system.

All tests were performed 5 times each, and the average of all test runs were used as the results.  Before the QuickBench and BlackMagic Speed Test, the SSD drives were cleared using Parted Magic's Secure Erase function and the hard drives were formatted.

For the real world file copy tests and boot speed tests, an identical system image based on my personal OS X install was used.  Wifi was turned off, and bootup applications included atMonitor and Google Chromium.

The hard drives and SSDs were connected to the SATA 6G connector in the optical bay using the OWC Data Doubler.

Other Testing Notes

For fun and because we had two of Kingston's HyperX SSDs on hand, we tested the drives in a RAID-0 configuration. RAID was setup in OS X's Disk Utility with a 128k stripe size, and the drives were installed in the main hard drive bay and the optical drive bay (using OWC's Data Doubler).


Performance Test Results

QuickBench helps to illustrate the maximum potential speeds that drives can achieve.  It uses easily compressible streams of data, so its results should be read as a best case scenario.  For the read tests, the all of the SSDs show similar performance.  Crucial's m4 is slightly slower for 4k random reads, and OWC's Electra 6G (representing the async Sandforce class of drives) is a bit slower for sequential reads.

For writes, the two synchronous NAND powered Sandforce drives from Kingston and OCZ lead.  Closely following is OWC's Electra 6G with asynchronous NAND.  Crucial's m4 was the slowest SSD in the test, but still far ahead of the two hard drives.

The BlackMagic Disk Speed Test uses incompressible data streams for testing drive performance.  As such, the performance of Crucial's m4 drive was relatively the same between the BlackMagic and QuickBench results.  On the other hand, the SandForce SF-2281 based drives benefit from compressible data, and their speed drops on the BlackMagic test versus QuickBench.  The Electra 6G drive suffers the most, due to it using asynchronous NAND.

Regardless of which SSD you choose, your boot times will be cut significantly.  While not specifically tested, application load times are generally also a fraction of what they would be with a hard drive.  All of the SSDs are within a second of each other.

Again, the HyperX and Vertex 3 drives lead the pack with regards to file copying speeds.  The m4 and Electra 6G trade blows between the tests.  The m4 wins with the Movie and Music tests as those files are not easily compressible, but the Electra 6G wins for the Document test as those files are more easily compressible.

In reference to our original Solid State Drive Roundup featuring these same drives but on a Windows platform, the SSDs performed as expected with OS X running on our MacBook Pro.  In general, the Sandforce SF-2281 drives with synchronous NAND (HyperX and Vertex 3) performed the best.  The Sandforce drive (Electra 6G) with asynchronous NAND and the m4 (with its Marvell controller) traded blows, dependant on the compressibility of the data used to test with.