Comments on: You Need A Cloud Based Grid Supercomputing Cluster http://www.linux-mag.com/id/7593/ Open Source, Open Standards Sat, 05 Oct 2013 13:48:18 +0000 hourly 1 http://wordpress.org/?v=3.1 By: gcreager http://www.linux-mag.com/id/7593/#comment-7223 gcreager Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7593/#comment-7223 <p>Doug: Good summation. I agree with virtually the whole article (and that\'s not always the case). I\'m getting tired of PHBs and marketing droids telling me I have to develop a grid... er... cloud infrastructure (or make Hadoop work under torque) because it\'s the latest hype. I come from the old school, where high performance computing meant eaking out the best performance from the hardware and interconnect fabric one could, not seeing how many different sites could be engaged in an MPI-fest with high latency, which in my mind, tends to describe today\'s cloud (and yesterday\'s grid) efforts.</p> <p>Nice job. </p> Doug: Good summation. I agree with virtually the whole article (and that\’s not always the case). I\’m getting tired of PHBs and marketing droids telling me I have to develop a grid… er… cloud infrastructure (or make Hadoop work under torque) because it\’s the latest hype. I come from the old school, where high performance computing meant eaking out the best performance from the hardware and interconnect fabric one could, not seeing how many different sites could be engaged in an MPI-fest with high latency, which in my mind, tends to describe today\’s cloud (and yesterday\’s grid) efforts.

Nice job.

]]>