@Carl Love Thank you very much. Awesome, you helped me a lot, really. I was thinking about the same problem with multiple indexes. Here is a simplified example of what I found.

for i from 1 to 9 do

for j from 1 to 9 do

for k  from 1 to 9 do

cat(A,i,j,k) :=ln(100*i+10*j +k) ;
#here a single number, but the same can be done with matrixes : 

save cat(A,i,j,k), sprintf("A%d.m",100*i+10*j+k);  

od

od

od 

May I ask if you would do it differently?

I really apreciated your help. Best,

@Carl Love 

Thanks a lot for looking at my question,

Actually I can’t do this because I do not have enough RAM in order to ask Maple to store all the data into a single matrix (even with 320 GB). Hence the necessity to build “sub-matrixes”: I can save part of the data and then free some RAM.

I hope this helps. Thank you in advance.