Got FLOPS? IBM does and you can use them……..for a fee.
The economics of supercomputing power and usage continue to evolve. Utility and on-demand computing aren’t new concepts, but for the first time IBM is making a large chunk of computing capacity available to its customers via its Blue Gene supercomputer in Rochester, NY. It’s unclear how well this cost recovery model will work for IBM using this machine, but it further demonstrates that there is a market for serious computational power without the need to own it.







March 25th, 2005 at 10:18 am
IBM is NOT a non-profit firm. Their fee schedule for services and roducts can often place their offerings far outside the budgetary constraints of start-ups, nonprofits, and research firms that could really benefit from gaining access to, in this case, some HPC.
Would they, IBM, be willing to make a deal?
In for-profit ventures/compute experimentation or modelling efforts, allocate time and receive payment in the form of a percentage? Yes, sign all those legal agreements, assess the venture, provide sufficient ‘How-to’ guidance to limit waste and insure proper use of the compute facility based upon stated goals, etc., etc., etc…
IBM is the leading big-money supporter of Open Source, they have really very little to lose in such an effort.
In non-profit ventures, well, put some of the money they have made on ‘for-profit’ activities back into funding true charitable needs for HPC. This would be just great science and a benefit for humanity in the arts and science/basic research and advancement of knowledge and make it known from the outset that taht is what this particular area MUST BE ABOUT!
It is ‘okay’ to have the six of the top ten supercompuers on the TOP 500 list, the majority all doing weapons and technology research for BIG-BROTHER, but we would have to do a lot less research on the next super-weapon if we ‘little people’ could get some time on those machines making effective vaccines, designing better high altitude low-eart-orbit net access satellites and gliders, or finding a preventative for the Guinea worm, curbing rampant biosphere degradation, etc., etc., etc… Right now cost and poor allocation along with RED TAPE is killng our hopes of EVER getting any HPC time!
I am listening,…