Skip to content

Conversation

@twoeths
Copy link
Contributor

@twoeths twoeths commented Sep 30, 2025

Motivation

  • for a common x86_64 server in our infrastructure, it's 1.8x faster to batch hash so we should find a way to utilize it
  • for Mac Silicon, batch hash is not really faster (see [hashtree-z] batch hash performance in aarch64 hashtree-z#7). However this implementation shows it's 10% faster than recursive hash consistently so we should use batch hash too

Descriptions

  • new batch hash api in pool
    • use a lazy_nodes ArrayList to store nodes at different levels to hash bottom-up
    • track start index of each level in a stack-allocated array and populate it along the way

Base automatically changed from te/fix_set_nodes_path_parents_alloc to main October 1, 2025 20:17
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants