So that's probably a processor refresh with associated PCI bandwidth improvement then. Internally there isn't a lot to block throughput in a Violin so these are the only real bottlenecks.
Violin Memory's flash array fightback starts with a present from the Easter bunny: a whole new array. Violin, you will recall, lost CEO Don Basile after its IPO was followed by deepening losses and falling share prices. Incoming CEO Kevin DeNuccio has stabilised the ship, refocussed engineering and recruited new executives, …
They still have software overhead and, if they keep building storage functionality, that will probably increase. IBM's FlashSystem is as close to straight bare metal as it gets, 100 microseconds of r/w latency. Not to say that software functionality is necessarily a bad thing, if that software functionality is useful, but the idea of all-Flash is to keep things as close to wire speed as possible. When companies, especially extreme performance storage companies, start building out every bell and whistle, it defeats the purpose.
speeds and feeds always win the sales battle
Said No One Evar.
Damage is already done for Violin. Someone get a fork, they are done.
Re: speeds and feeds always win the sales battle
Agree, now that the major storage providers, EMC and IBM, have their own all-Flash arrays and HP is going to plod along with their SSD = Flash array (which it doesn't) strategy, I don't see much room for Violin. Their best bet was that they would get acquired by HP, but that isn't going to happen as HP just spent billions on 3PAR.