receiving, using at least one processor, a query associated with an index join over a block of columns, the query accessing data stored in a database, the database storing data in a plurality of tables, each table having a plurality of columns and a plurality of rows;

in response to receiving the query, identifying, using the at least one processor, one or more join values, the one or more join values corresponding to the block of columns for the index join, wherein the identifying further comprises identifying one or more unique join values in the one or more join values and using the one or more unique join values to identify one or more value identifiers corresponding to the identified one or more unique join values;

determining, using the at least one processor, one or more count values associated with each of the one or more value identifiers;

determining, using the at least one processor, one or more row positions in the plurality of rows corresponding to the determined one or more count values and the one or more value identifiers;

generating, using the at least one processor, a mapping among the one or more count values, the one or more value identifiers, and the one or more row positions;

executing, using the at least one processor and based on the mapping, a join of the one or more join values based on the rows including the one or more join values; and

outputting, using the at least one processor, the joined one or more values.