There is a similar question asked and answered here: K-Means Cluster with R
Check it out, it may help.
There are 2 suggestions in it:
1. (Marked as answer) It was suggested that the table calculation was sending data to R one row at a time - check it out and correct if this is the case
2. Using result <- kmeans(x = data.frame(.arg1,.arg2,.arg3), 1, iter.max = 10