Xbench Disk Comparison

Using Xbench, I compared the three disks attached to my Powermac.

First, I have the stock drive, a Deskstar 120GXP (80 GB).

Disk Test   100.77  
    Sequential  103.00  
        Uncached Write  90.65   37.79 MB/sec [4K blocks]
        Uncached Write  90.87   37.21 MB/sec [256K blocks]
        Uncached Read   170.74  27.03 MB/sec [4K blocks]
        Uncached Read   91.39   36.93 MB/sec [256K blocks]
    Random  98.64   
        Uncached Write  101.96  1.53 MB/sec [4K blocks]
        Uncached Write  100.45  22.66 MB/sec [256K blocks]
        Uncached Read   95.93   0.63 MB/sec [4K blocks]
        Uncached Read   96.47   19.85 MB/sec [256K blocks]

The next drive is a Maxtor DiamondMax Plus 9 (200GB) in an OWC Firewire enclosure.

Disk Test   86.58   
    Sequential  71.33   
        Uncached Write  76.91   32.06 MB/sec [4K blocks]
        Uncached Write  75.22   30.80 MB/sec [256K blocks]
        Uncached Read   55.75   8.82 MB/sec [4K blocks]
        Uncached Read   84.44   34.12 MB/sec [256K blocks]
    Random  110.12  
        Uncached Write  128.62  1.93 MB/sec [4K blocks]
        Uncached Write  135.84  30.64 MB/sec [256K blocks]
        Uncached Read   97.47   0.64 MB/sec [4K blocks]
        Uncached Read   91.51   18.83 MB/sec [256K blocks]

Finally, the drive I just installed is a Maxtor DiamondMax 10 SATA drive.

Disk Test   149.91  
    Sequential  159.95  
        Uncached Write  176.23  73.46 MB/sec [4K blocks]
        Uncached Write  172.24  70.53 MB/sec [256K blocks]
        Uncached Read   143.74  22.76 MB/sec [4K blocks]
        Uncached Read   152.20  61.49 MB/sec [256K blocks]
    Random  141.05  
        Uncached Write  744.83  11.17 MB/sec [4K blocks]
        Uncached Write  141.13  31.83 MB/sec [256K blocks]
        Uncached Read   108.76  0.72 MB/sec [4K blocks]
        Uncached Read   93.14   19.17 MB/sec [256K blocks]

The new drive is by far the best. In particular, the first random uncached write speed is amazing. However, it was a little slower than the stock drive by just a little bit. This is a little perplexing, but considering that Xbench generally isn’t used as a disk benchmarking utility, maybe the results aren’t perfect. Unfortunately for all of you, I’m not willing to spend the $25 for QuickBench, so you’ll have to refer to Barefeats.