Microsoft Excel Accelerator Using Linux Cluster/Grid

[anonxyz]

Hi
There is a strange program from Aspeed called Accelerant that can increase the speed of Microsoft Excel Spreadsheet application using Linux based Cluster /Grid.

Does this software really works?
Is there any independent benchmark of this program?

Is it possible to use the optimizer on other Microsoft applications?
Is there any other similar Dual/Quad core optimizer program?

From FAQ on their website:
"We already use a SOA approach to applications, where does ASPEED fit?
ASPEED is simply a SOA performance expansion. It enables you to easily run applications with parallelized components as part of the application architecture. For example an Excel program can actually be doing its parallelized calculations on a Linux cluster or grid. The same calculations can also be used by runs submitted as batch runs directly to the Linux cluster."

[bigtomrodney]

While I don't doubt this is possible, I am always concerned when I see Office based solutions that are important enough or hungry enough to need this sort of solution. I regularly come across people where I work who use Excel like databases - surely a better solution is to run a SQL database? I've heard the argument that the users implementing the solution don't have such experience to be qualified to do so, but that usually begs the question "Why are you doing something so critical without either a suitable platform or without suitable experience?"

I'm sorry if I sound cynical, I genuinely apologise for that. I deal with data on a daily basis in a kind of grey area between IT support and End-User. I am a data analyst. I am constantly pulled in as a secondary task to try and 'fix' problems with systems that were under-engineered, such as Excel macros designed to deal with 800 staff members' personal data and pay - though not for a month, but rather for the past 3 years. Much the same where access is used to keep mission critical data and when the database becomes corrupt the blame is pointed outwards instead of looking at the source of the problem.

Would a better solution be to implement a basic SQL platform combined with reporting tools?