There was a nice article from the HPC perspective a while ago, that discusses why Hadoop didn’t take off in HPC:
http://glennklockwood.blogspot.de/2014/05/hadoops-uncomfortable-fit-in-hpc.html
HPC with MPI and tools like Spark are for fundamentally different types of computational problems.
There was a nice article from the HPC perspective a while ago, that discusses why Hadoop didn’t take off in HPC:
http://glennklockwood.blogspot.de/2014/05/hadoops-uncomfortable-fit-in-hpc.html
HPC with MPI and tools like Spark are for fundamentally different types of computational problems.