Terascala appears to have been disassembled, with much of the team going to Cray. Terascala started out selling internally developed storage appliances for Lustre. They developed deployment, monitoring, and management tools. Their UI was reasonably good.
Then they struck up a deal with Dell and a few others. In doing so, they largely stopped their appliance sales. Put their code upon their partners hardware. This did generate more force multipliers for them in sales, but it cost them some of their differentiation … unless their boxes were entirely undifferentiated, where it would reduce their overall costs to avoid selling undifferentiated hardware.
Aside: We believe very strongly in highly differentiated hardware stacks as well as good software stacks. You’ll see a white paper coming out soon which shows what differentiated hardware can do relative to “cheap stuff”.
This gets back to their differentiation. They had a UI, tools, monitoring and reporting. They didn’t control Lustre, Whamcloud effectively did (and now Intel does). Which meant that they were effectively a management layer atop Lustre, which Whamcloud were, themselves, trying to create pre-acquisition from Intel.
This is the classical definition of being between a rock and a hard place. They didn’t control much of their tech that they depended upon for their product, and only offered a small differentiated UI/tools/monitoring/management system.
I think their play was for an acquisition by Dell.
From folks I’ve spoken with elsewhere, this is a dangerous strategy. One not likely to pay off for a smaller firm.
The following is speculation.
They raised capital a while ago, and I am guessing that the money ran out, and they couldn’t raise another round. So faced with closing down, they found someone to pull their team over and work on similar things, while the remaining customers on Terascala will be supported via a much smaller organization going forward.
Indeed, if you read the PR, you get a sense of what happened.
They didn’t acquire tech, just people, and not all of them. It is sad that they ended this way, but I don’t generally expect non-differentiated solutions, or pure software solutions (see CoRaid for examples of non-differentiated) to survive.