For some people mathematical projections aren't good enough and you/they need to see it in action. I guess I can appreciate that.
That's because thunderbolt 1 (which is basically a 1 channel version of TB 2) already fails to get up to maximum speeds.
Thunderbolt RAID 0 with 6Gb/s SSDs appears to run into a bottleneck when you compare it to the SAS RAID 0 with the same 6Gb/s SSDs. I guess the 1000+MB/s theoretical bandwidth is... theoretical.-
Barefeats
Whereas most Thunderbolt storage devices top out at 800 - 900MB/s, Thunderbolt 2 should raise that to around 1500MB/s (overhead and PCIe limits will stop you from getting anywhere near the max spec).
-
AnandTech
Also, the bandwidth limit is a lot more than you think. The nMP, according to deconstruct60, is only 6 channels--60Gbits/sec. That means that running a couple 1.5GBps cards in a RAID (ignoring the insane bottleneck of software RAID you'd have to suffer) is going to take 2/3 of the bandwidth.
As far as 36 devices, you got me there, that's technically more than PCIe can directly plug in...indirectly, however, PCIe can run controller cards for fire-wire, SAS, SATA, USB, etc which can bring that number up in the hundreds. Is it as much as thunderbolt? No? Do you need more than that? Probably not. At that point you should probably be concerned about bandwidth....
That's not to mention that NONE of your 36 little thunderbolt-connected devices can be a 3,000MBps hard drive, a GPU (unless you'd like
some pretty great throttling), or a 4 port SAS controller. TB2 is 1/8 the speed of PCIe 3.0 16x.
----------
Meanwhile I did my time, I got my CS masters, and I've been working in a related field for the past 30 years - so I can just use the math to formulate a pretty good estimation.
Holy Condescending tone, batman. You just found out the product you've been touting for months is literally half as fast as you've been assuming and I'm the uneducated one? Plus you're poo-pooing experimental data when the tech you're talking about has already proven to have practical maximums well below the theoretical.
I guess we're all supposed to digest the totally false information you're putting out there just because you got a piece of paper?
You're wrong, Elenore!
Did you know CS 30 years ago has almost nothing to do with anything today? I also like how you assume I don't have a degree in computers as well (I have 2).