"Typical server configuration for running Rapidminer, for practical usage?"

ruser

I have been using the Rapidminer on my PC (2 GB RAM, 2 GHz Dual core CPU) for the last few months. What I observed was that the process takes the complete resources in my PC. It is obvious that when using Data mining, we'll have to deal with huge amount of data (maybe, few tarabytes, for example) lying in the Datawarehouse.
I have also realized now that I should no more run the Rapidminer process in my low-end PC, and I'll have to start using it on a system with higher configutration. How do I find the approximate CPU, memory configuration that I typically require to run the Rapidminer process for such a huge DB as the input data source? What do you guys normally use, as the server for running Rapidminer?

Advance thanks for your inputs!

haddock

http://rapid-i.com/rapidforum/index.php/topic,537.msg2053.html#msg2053

Roberto

ruser,

I can tell you from experience that if you have the parallel validation operators they GREATLY increase the performance of your process. In our case, we are evaluating huge amounts of compiled biological data (27,580 attributes per example, upwards of 60 samples to train on).  In addition to the enterprise edition of rapidminer, we also built a miniserver to help speed things up as well, as the datasets are only going to get larger if this works the way we want it to. It has 48GB 1333Mhz DDR3 ram, asus motherboard with dual quad i7 (2.27Ghz). It may be a little bit overkill, as we went a little crazy because we had some new grants get funded, but it has worked like a charm in speeding up the process for us.  So if your group can afford the enterprise then go all out with it I would say.  Total cost for everything for us was about 6500 dollars (US) which includes the computer + feature selection plugin + rapid i ee small enterprise edition (academic version)