↝ OWC / MacSales.com... ↜
↝ diglloyd Deal Finder... ↜
Buy other stuff at Amazon.com...
Work from home?
Consult with Lloyd before buying!
System configuration for CPU, GPU, memory, SSD, backup, RAID, optimizing performance, workflow.
For YOUR optimal workflow
Avoid costly mistakes in throwing money at the problem for less performance and/or buying the wrong machine for your own needs!
Remote screen sharing available for RAID, software configuration, etc.
More about 2019 Mac Pro vs iMac 5K vs iMac Pro...
RAID-0, RAID-5, RAID 1+0 Performance (SoftRAID, Thunderbolt v1)
Related: hard drive, Other World Computing, OWC Thunderbay, RAID, RAID-0, RAID-4, RAID-5, SoftRAID, storage, Thunderbolt, video
Tests shown here uses the fastest portion of the drives in order to show the peak speeds possible with the Thunderbay; all hard drives slow down as they fill up, as per the simple physical fact of πD (track circumference).
The Thunderbay performed admirably, as did the Toshiba MG03ACA300 hard drives, which offer very impressive performance.
Tested with the Thunderbay IV (Thunderbolt, not Thunderbolt 2).
Results shown are real-world throughput through the file system API (not unachievable driver-level throughput).
The performance is outstanding—using four hard drives, these speeds are exceptional. It demonstrates not only that the drives are fast but that the Thunderbay itself is an efficient unit that does not get in the way of performance. Consider that four solid state drives top out not much beyond this level in the Pegasus J4; the speed approaches Thunderbolt itself as a limiting factor.
RAID 0: 720 / 753 write/read MB/sec RAID 5: 534 / 563 write/read MB/sec RAID 1+0: 364 / 375 write/read MB/sec
Note that RAID 5 (striping with parity) and RAID 1+0 (striped pair of mirrors) offer fault tolerance, yet the performance level remains in excess of what most programs can benefit from or take advantage of. Even Photoshop tops out around 500-600 MB/sec before running into its own internal bottlenecks when saving or opening big files.