Excellent blog by Guident – http://tinyurl.com/bne6t8k
comparing a Hadoop versus Traditional High Performance Computing. A specific use case of reading large log files is compared and Hadoop is the winner in terms of performance.
What happens if we have access to traditional HPC hardware? Should we use Hadoop on HPC? Check out excellent article by S.Krishnan on this – http://tinyurl.com/cwbzvof . Results are not conclusive, but an interesting read.
Bottom line, it appears to be dependent on the use case? Has anyone done a more detailed comparison?