I am using
A <- as.matrix(read.table(fname))
A <- as.matrix(fread(fname))
As Gregor mentioned, fread has an argument "header" which defaults to true if all of the entries in the first row are character type fields. When you ran fread it most likely assumed the first row of data contained column names and the remaining 999 rows were data. I couldn't say for sure without seeing the data you're importing, though.
This should do the trick for you:
A <- as.matrix(fread(fname, header = F))