In the last case they were like 100x100, but only two rows or columns were not zero, so only 2% full. We were using einsums to multiply them with 4D arrays, and I solved the problem writing some code with @guvectorize.

In al old case they like 1000x1000, like a block checkboard where one of the colors had only 0, so like 50% full, but the blocks had different sizes. It was an old project in Fortran, and we used a "for"(do) for the blocks and another inside each block.

Neat. That’s fairly sparse. I’m 0% surprised to hear that small sparse matrices haven’t got as existing code out there, seems like a good excuse to roll your own :)