Hadoop just runs as a standard java process, you should find something that bridges between OpenCL and java, a quick google search yields: http://www.jocl.org/
I expect that you'll find everything you need to accomplish the handoff from your mapreduce code to OpenCL there. As for HDFS, hadoop will generally handle marshaling data in/out of HDFS, remember that you're thinking of your problem in terms of KEY,VALUE pairs, you're going to implement a map(.) and reduce(.) method and in those methods you'll pass the data to/from OpenCL via the OpenCL java bindings. It's quite common to need to run multiple map/reduce steps to accomplish an end goal. David From: rohit sarewar [mailto:rohitsare...@gmail.com] Sent: Monday, May 13, 2013 8:35 PM To: user@hadoop.apache.org Subject: Access HDFS from OpenCL Hi All My data set resides in HDFS. I need to compute 5 metrics, among which 2 are compute intensive. So I want to compute those 2 metrics on GPU using OpenCL and the rest 3 metrics using java map reduce code on Hadoop. How can I pass data from HDFS to GPU ? or How can my opencl code access data from HDFS ? How can I trigger OpenCL codes from my Java map reduce codes ? It would be great if someone could share a sample code. Thanks& Regards Rohit Sarewar