Skip to main content
Fig. 2 | Source Code for Biology and Medicine

Fig. 2

From: FlexDM: Simple, parallel and fault-tolerant data mining using WEKA

Fig. 2

FlexDM parallel processing speed-up. Relative execution time for a FlexDM-enhanced WEKA experiment as a function of the number of allocated CPU cores (a user-specified parameter which defaults to the number of available cores minus 1). This test was completed on a desktop PC with a quad-core Intel i7 processor. A near-linear speed-up is evident for n = 1:4 cores, with divergence due to background system tasks necessary within the host OS environment. For n = 5:8 cores, the i7 processor’s hyper-threading capability is leveraged for further quasi-linear speed-up, resulting in an overall 5-fold reduction in execution time when distributed across 8 logical CPU cores

Back to article page