fcpx.at

Corning Optical Thunderbolt Cable

If you are like me and prefer to edit in a silent environment, then maybe you habe thought about banning noisy hardware from your edit suite into the adjacent room or in a central machine room, if you're in a larger facility.

I wanted to move the noisy Promise Raid to the adjacent room to keep my edit bay quiet. For a while I did that using a 2 Meter Apple Thunderbolt cable, which restricted the placement of both the MacPro and the RAID next room. Now that I want to get one of those stand-to-sit desks, which meant that those 2 meters were not long enough.

I knew about the 10m optical Thunderbolt cable that Corning offered since I switched to Thunderbolt periphery, but the price tag of around €299.- was too steep for me. By accident I stumbled across a newer 5m cable on offer in the Apple store that I aquired for €179. Still a pricey bit of equipment, considering it's a goddamn cable, but a sum I was prepared to pay, considering the health benefits from that stand-to-sit desk.

corning

But let's move on to the numbers.


Test Environment

  • 2013 MacPro 8 Core, 64GB Ram, OSX 10.10.3, Dual D700
  • Corning 5m Optical Thunderbolt Cable
  • Apple 2m Thunderbolt Cable
  • Promise R4 Thunderbolt 2 RAID (Raid 5 configuration)
  • Blackmagicdesign Ultra Studio 3D (Thunderbolt 1 only)
  • Blackmagic Speed Test utility (set to 2GB file size)

The first test I did was measuring the speed of my then current configuration: The Promise-RAID directly connected with an 2m Apple Thunderbolt cable. What I got were write speeds of around 550 MB/sec and read speeds of around 400MB/sec — pretty much what I expected. Remember: the Promise R4 is RAID5 configured, so that one drive can fail. I am willing to sacrifice the speed of RAID1 for that extra bit of safety, because I rarely ever deal with more than HD resolution ProRes files, for which the RAID5 speed is plenty.
One thing I don't quite get though is why in RAIDs WRITE is faster than READ. Should it not be the opposite? I am wondering about that since I got my first RAID 10 years ago — maybe one of you can enlighten me on that.

Test Apple TB Cable


The next Test I did was to replace the Apple TB cable with the Corning one. There was no speed change whatsoever. The little x in the table 1080p50/Read als disappeared a second after I took the screenshot below.

Corning TB Cable

With my current configuration I also had the Blackmagic Ultra Studio 3D connected directly to the MacPro with its own Thunderbolt cable. Now that was not possible, because I would have needed another one of those corning cables, so I decided to daisy-chain the Ultra Studio (Thunderbolt 1) after the Promise Raid (Thunderbolt 2). Remember that when daisy chaining TB1 and TB2 devices together to have the faster device closer to the host computer, everything connected after a TB1 device will only get TB1 speed regardless of it's TB capabilities.
The results are interesting:

 Test Corning TB + BM UltraStudio 3D

Now Blackmagic Speed test showed transfer rates of 930 MB/sec write and 900 MB/sec read! And not only was the Blackmagic Ultra Studio 3D connected, but I also had Final Cut Pro X play back a ProResHQ file (from the SSD system drive). I was surprised to see that kind of performance with the two devices daisy-chained.
Can someone explain to me why this configuration ist faster than when each device gets it's own port on the MacPro?

So the conclusion is that I don't know why is is the way it is, but the Corning cable works — way better than expected. Nothing stands in the way of my sit-to-stand desk now.

— Florian


Back

Archive