AVL Tree vs Balanced Tree

Question

AVL tree is a well known internal memory data structure whereas Balanced tree is known external memory data structure. Why can't we use Balanced trees for internal memory also?

Answer 1

You very much can have balanced trees in internal memory. An AVL tree is simply one type of balanced tree and there are others as well, such as red-black and 2-3-4 trees.

So, I'm not sure where you got the idea balanced trees couldn't exist in internal memory but I'd be rethinking that if I were you.

And, in fact, you could put an AVL tree on disk as well if you were so inclined.

---

I suspect, based on your comments, that what you may be thinking of is a BTree which is like a binary tree but each node can hold more than one value and have more than two children, such as:

root node -,
           |
           V
 +------+------+------+
 | Val1 | Val2 | val3 |
 +------+------+------+
/       |      |      \
<other nodes down here>

That's different to the more general term 'balanced tree' and they're often used in disk situations as you tend to want to read/write entire blocks/clusters/sectors at a time.

So, if you can fit ten values into a disk block, it's more efficient to use a BTree (whereas memory does not have a concept of block size so it can be preferable to use a simpler algorithm - BTrees have to combine both tree search to find a node and linear/binary search to locate the value in the node).

But, while BTrees can be a type of balanced tree, it's again only one type.