RAID 0 (Striped) Intel X25-M 160GB SSD drives, MacBook Pro 17" unibody

After three years on my old MacBook Pro 15", I purchased a new MacBook Pro 17" unibody. It's gorgeous. I elected to get the glossy screen because I often work outside or in areas with bright light streaming in from windows (e.g. boat salons). Matte screens simply don't work in those environments. Go ahead -- tell me I'm wrong, and I'll ignore you as I continue to be productive. Anyway, continuing on my path to create the ultimate mobile computing platform (prev: RAID 0 in MBP, OCZ RAID 0 SSD in MBP), I shoved two Intel X25-M 160GB SSD into my new MBP 17", again, using the MCE Tech Optibay. On these newer MacBook Pros, the Superdrive is (finally) SATA. I benchmarked the Intel X25-M SSD in both the standard internal SATA bay as well as in the SATA Optibay, and there was no performance hit like there was with the old Optibay, which used a IDE-SATA bridge.

So, why did I purchase the Intel X25-M over another drive, like the OCZ Vertex, which has also been favorably reviewed?

I did so because I wanted a SSD that would perform well while multitasking. Last time, I made the mistake of purchasing OCZ Core Series v2 drives, which use the notorious JMicron MLC controller, resulting in terrible write delays. On that system, the system screamed along just fine if I was only doing one task at a time, but as soon as I started doing any sort of serious multitasking, it punished me by slowing down to a crawl.

If you're interested in going SSD, you should absolutely read this article: The SSD Anthology: Understanding SSDs and New Drives from OCZ. It's the best article I've read that compares SSDs that are just hitting the market (and, it tells an interesting story about OCZ and their philosophy as a company). I've also read that the Intel SSDs run cool to the touch. After using the old MacBook Pro and hot-to-the-touch OCZ Core Series v2 drives, I've had enough crotch burning.

I chose the Intel.

Before I did anything with the new drives, I updated their firmware from 045C8790 to 045C8820. Here's a useful article on the firmware update, and what it does: Intel Responds to Fragmentation with New X25-M Firmware.

I'm sorry I didn't take any photos of the system with two SSDs in it. I'm sure there are some already out there.

The results are incredible, and when switching from application to application, the system doesn't pause at all. In fact, I can almost never tell that there is something going on in the background -- that is, assuming that the operations are disk-bound and not CPU-bound. Restoring a VMWare Fusion virtual machine happens in a matter of seconds, even when it hasn't yet been cached. When it's in the cache, Fusion barely has time to display the drop-down status flap before my Windows instance is ready to go.

I've finally found the right combination! Hopefully, it will take me another three years into the future before I have to think about upgrading again.

BENCHMARKS (QuickBench) MacBook Pro 17" Unibody 2.93Ghz Intel Core 2 Duo 2nd internal SATA drive installed using MCE Technologies OptiBay Hard Drive adapter Mac OS 10.5.6, Apple software RAID

Striped (RAID 0) Intel X25-M Sequential Read Benchmarks

Striped (RAID 0) Intel X25-M Sequential Write Benchmarks

Striped (RAID 0) Intel X25-M Random Read Benchmarks

Striped (RAID 0) Intel X25-M Random Write Benchmarks

BENCHMARKS: THE NUMBERS (QuickBench) All values are in MB/s

INTEL X25-M SATA 2 x INTEL X25-M SATA RAID 0
Size Seq. Read Seq. Write Ran. Read Ran. Write     Size Seq. Read Seq. Write Ran. Read Ran. Write
4 KB 31.85 25.55 10.10 27.14 4 KB 28.60 23.78 9.52 25.06
8 KB 57.54 30.49 19.58 46.35 8 KB 51.91 41.43 19.49 43.13
16 KB 93.41 60.70 35.33 62.49 16 KB 86.27 76.89 38.06 63.65
32 KB 113.75 63.35 62.08 64.50 32 KB 141.27 97.61 62.28 80.15
64 KB 132.60 68.56 99.49 70.10 64 KB 188.05 97.93 110.82 81.65
128 KB 178.26 74.23 144.01 76.17 128 KB 265.47 117.16 187.60 112.45
256 KB 211.41 77.46 186.80 78.85 256 KB 337.05 122.75 266.91 117.40
512 KB 236.13 79.04 218.15 79.94 512 KB 398.48 135.09 343.67 128.12
1024 KB 248.80 80.12 239.27 80.41 1024 KB 451.86 141.95 415.26 138.33
2 MB 254.97 82.38 2 MB 461.27 145.69
3 MB 263.73 79.04 3 MB 476.10 151.36
4 MB 262.89 83.05 4 MB 482.10 149.18
5 MB 263.75 79.99 5 MB 485.78 153.15
6 MB 264.57 80.01 6 MB 489.77 150.37
7 MB 265.24 81.10 7 MB 496.41 153.10
8 MB 265.20 78.80 8 MB 498.16 155.70
9 MB 266.37 80.79 9 MB 505.32 143 .93
10 MB 263.47 79.25 10 MB 495.44 145.21
20 MB 268.29 78.04 20 MB 518.20 152.36
30 MB 267.82 78.10 30 MB 525.52 144.51
40 MB 265.52 77.35 40 MB 526.34 143.84
50 MB 264.76 75.66 50 MB 525.06 148.06
60 MB 267.71 77.93 60 MB 526.03 149.64
70 MB 267.37 77.96 70 MB 531.73 153.58
80 MB 267.08 74.93 80 MB 531.52 153.16
90 MB 267.61 74.99 90 MB 525.44 141.02
100 MB 268.60 75.32 100 MB 529.91 124.03

XBENCH RESULTS

I also ran Xbench on my system (booted fully configured -- no attempt to boot cleanly), and it scored a 239.41 [results as txt].

ADDENDUM

For those of you attempting this, be sure to check out how the new MBP SuperDrive differs from the one MCE Tech puts in their home installation instructions: MacBook Pro 17" Unibody SuperDrive area change

ANOTHER ADDENDUM

I got attacked on a forum for stating that the Vertex uses a Samsung controller (it doesn't). That was wrong, and I acknowledged my mistake. But I still am sticking with my decision to go with Intel. Look at this Anandtech article, for example. While the Vertex outperforms the Intel in sequential writes and reads, it gets its ass kicked in random reads and writes. Here's what the article concludes:

> My recommendation still stands: the Intel X25-M is by far the cream of the crop of the desktop SSD world. The Indilinx based drives have the potential to be good, lower cost alternatives to the Intel drive but you still have to approach them with caution. While the OCZ Vertex drive worked fine in my tests and on my testbed, this is a brand new drive with a controller from a company without a proven track record.

> In the coming weeks I will be looking at the latest updates to Samsung’s MLC controller. While I haven’t been terribly impressed with the performance of Samsung drives thus far, they are at least reliable and compatible.

> While I’ve heard that the JMicron based drives are no longer selling very well, it looks like at least a few manufacturers are going to be using the JMF602B controllers to deliver 512GB SSDs in the coming months. Buyer beware.

> I will say this: outside of Intel, Indilinx appears to know what is important when it comes to SSD performance. With more companies releasing drives based on the Barefoot controller, we should hopefully see any compatibility problems get sorted out faster.

> There is some even better news that has surfaced since last week. If the 1275 revision ends up being problem-free, it does deliver more than 3x the random-write performance of the Vertex I first previewed and a more than noticeable 10% boost in application performance. That’s not enough to dethrone Intel, but it is more than enough to make the Vertex even more desirable.

> The gotcha still applies: this is the first version of an Indilinx controller we’ve seen in the desktop market. Compatibility and longevity have yet to be proven; so far my experience has been positive but that’s merely one datapoint. We have a long way to go my friends.