Hello,

I have recently purchased a G-Speed es 4TB with G-Tech raid controller. The benchmark test from AJA shows as 195MB/s write & 230MB/s read. However, when I load up FCP, there does not seem to be any performance boost anywhere in either playback or Render speed over the single 1TB drive i was using previous, even though the read/write speed on tha twas 100MB/s. I have moved my scratch disk to the G-Speed and it is in RAID 0 using factory defaults.

Are there any settings outside of the G-Speed web controller setup (either in OSX or FCP)? Or is $1000 enough to spend to notice a speed enhancement?


Any help would be appreciated.

Thanks

100MBps is pretty decent performance from your original disk & as you see no improvement from a disk subsystem with double the performance then it would indicate that in this case the disk is not the bottleneck. It depends how complex your project is e.g. how many video streams etc as to the I/O requirements.

A while ago I installed an eSATA RAID 0 array on my wife's Mac Pro but like you we saw no noticeable improvement in FCP performance so ended up removing the RAID array as it was noisy.

If you have a $1000 to spend on souping up your Mac Pro then investing your money in a couple of Solid State Disks will pay dividends & leave a few hundred change. Use one SSD for OS X & applications & the other as your working disk. SSDs are many times faster then regular hard disks & the price for large capacity SSDs has reduced a lot. They are way more reliable too.

As long as performance you may not feel it, you can take advantage with your G-Raid with multi-cam editing.

Have you tried playing multiple streams of video?

Have you tried using Compressor?

These are things where I notice a big improvement with a RAID. (we had the same RAID as yours)

Mark,

i'm not familiar with the G-tech RAID controller card however there are most likely two different ways to have the RAID physically created:

One is to have the CARD create and manage the array. If that's what was done then this might be where the data bottleneck lies, as most eSATA cards do not have the internal bandwidth to create AND manage the array in real-time very fast.

Or, you could create a *software* RAID-0 in Mac OSX, which is done using Disk Utility. That's where each drive from the RAID enclosure is actually in JBOD mode and your OS creates the array in software. If you use this method you should see exactly the same speed-test benchmarks that are advertised.

Try creating a software-RAID and see what happens and let us know.

Robert, there is no disk bottleneck. The G-Speed es 4TB with G-Tech raid controller is performing pretty much to what G-Tech claim as per this screenshot from Aja benchmark on their website with Write 207.5MB/s Read 229.8MB/s for a four member RAID-0 array

Mark's problem is that he sees no increased performance even though he is using this RAID array & the reason for that as I explained in my previous reply is that he did not have a disk bottleneck in the first place. As others have noted with more complex multicam projects he should see performance gains.

G-SPEED eS - External RAID and JBOD Up to 12TB | G-Technology (http://g-technology.com/products/g-speed-es.cfm)