HADOOP x SPARC/Solaris
I've been hearing about many "exciting" news on tackling big data.Grid computation is proliferating on top of commodity hardware and Linux and HADOOP seems to be one of the latest variation.
Well, I have nothing against grid computing but as long as each node is scaled-up to its full potential before starting to scale-out.
Then you may have gotten my point. Yes, I think that SPARC/Solaris would be the ideal node for integral computation.
I regret that floating-point computation isn't present on any foreseeable future plans.
But back to HADOOP, I'm not convinced of its success in the long term.
Well, I have nothing against grid computing but as long as each node is scaled-up to its full potential before starting to scale-out.
Then you may have gotten my point. Yes, I think that SPARC/Solaris would be the ideal node for integral computation.
I regret that floating-point computation isn't present on any foreseeable future plans.
But back to HADOOP, I'm not convinced of its success in the long term.
0