>> Where as cloudcomputing is used for data coloboration and integrations among many users mainly.
But you can use cloud computing to accomplish some tasks which supercomputing does by using distributed processors in parallel which cloud offers you. I think they are related in how you can use them to accomplish tasks.
>> @goafrit nimbix offer their own cloud that includes Multicore processors, GPU's and Dsp's and i believe they are working on adding FPGA computing(using convey) for their system.
What do you mean Nimbix offers cloud that includes multicore processors? What is thar relevant to know the processors and GPUs in cloud since it is about your accessing it and not what they have used to host it. An explanation??
that is about $420 per processor - so about the price of a low end laptop (or an Ipad) for something with quite a lot more oomph. You can pay anywhere between $200 and $2000 just for a Xeon processor (admittedly in somewhat lower volumes!). I wonder what the spec of the processor's is.
And of course that much RAM, cabinet, disk and cabling etc does not come cheap either
Supercomputing and Cloudcomputing are two different terms leading to different technologies and requirements. Generaly heavy processing speeds algorithms needs supercomputers mailly used in researches at the universities except few exceptional cases. Where as cloudcomputing is used for data coloboration and integrations among many users mainly.
>> The Iridis4, which the University of Southampton ranks as one of the top 10 fastest supercomputers in the UK, was purchased from IBM for $5.1 million
That is a lot of money for a machine. No wonder IBM is building its future in this space if that is how much it costs. I think IBM should figure out a way to make that power available in the cloud so that people can rent easily. Not many institutions can afford that in this era of austerity. Imagine where I can plug in via the web and get Watson to work for me, say for 10 minutes, without breaking a bank.
A Book For All Reasons Bernard Cole3 comments Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...