Coachella Prep Day Two – SSD Testing
We were planning on making this our first Jam Day, but Brett was sidetracked early on with a logistics problem – running a few hundred feet of SDI cable to the DL3 Projectors that will be rigged up by the lighting crew. A few hours, hundreds of lines of small print, an undisclosed sum of money and a few million burnt brain cells later, we’ve got it all sorted.
Meanwhile, I got to borrow one of the SSDs (Solid State Disk) that they’ve been prepping, and run some tests of my own.
It feels like I’m always on the search for faster drives. I bought a 500GB external 7200RPM drive last year whose cases supported eSATA, which brought me a decent boost, although it did require buying an eSATA ExpressCard to be able to plug it into my MacBook Pro. When I heard that the Xochi folks had a couple of SSDs kicking around, my geek centers all lit up.
Extremely Unscientific Test Methods Resulting in a Fancy Graph
I used Disk Speed Bench X to run the tests. I quit all applications except for TextEdit and DiskSpeedBenchX, and the standard stuff I have running in the background – QuickSilver, Growl, Synergy, HardwareGrowler, probably some others. I ran the test about 10 times for each volume in order to get an average. My test system is a first-gen MacBook Pro Core Duo 2GHz with 2GB RAM, and a 100GB 7200rpm internal hard disk which is a little too full (about 5GB free).
The other drives tested are:
- LaCie Rugged FW400/800/USB enclosure with a Hitachi 100GB 7200RPM 2.5″ drive inside
- Cavalry CAXM37500 USB/eSATA enclosure with a Hitachi 500GB 7200RPM 3.5″ drive inside
- Cavalry CAXM37500 USB/eSATA enclosure with an OCZ Vertex Series 120GB 2.5″ SSD inside
I just switched out the drive in my Cavalry case to try out the SSD – I’ll get a mini eSATA enclosure for my own SSD when it arrives. I used a Bytecc 2-Port eSATA ExpressCard to connect the enclosure to my laptop (which has a Silicon Image 3132 SATALink controller for you info-junkies out there).
The SSD blew the other drives away in these raw read tests. We’re talking 2.6 times faster than my my internal or the firewire, and twice as zippy as my fastest platter-based disk.
The downside is that SSD is still quite expensive – this 120GB I was testing with is $345 at NewEgg. For the moment, I’m getting a 30GB drive – it’s teeny by today’s standards, but that’s okay – I can simply load one show’s worth of clips onto it, and keep my big externals for storing the full content libraries, etc.
DISCLAIMER: I know very little about testing hard drives. My simplistic methods are probably terribly flawed and my data is probably useless. But damn it is impressive how the drive performed in this particular benchmark application.
That said, I ran another even less-scientific test where I loaded up my Incisor AV Remix patch with all the media on the SSD. I did NOT preload the clips, and yet it ran just as well as it usually does after I preload entire clips into RAM from my external drive. This is huge, because it means I’m no longer limited by the amount of RAM in my machine, but by an upgradeable external device. Also, VJ Aquine feels much better running her Quartz Composer mixer from an SSD because she cannot preload her clips, and is therefore accessing the drive constantly, which could wear down a traditional platter-based system right quick.
As night fell, Brett continued wrangling cables in our Ghetto-Tech system and we did get to jam a bit. For this round, I stuck to vector-y foreground elements, with VJ Metamer mixing up the background, and running my foregrounds as a mask for various layers. I’m digging it.