parallel copy

Open
#4,540 0 comments 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
5/5
Estimated time
Over a week
Newbie friendliness
30/100
Issue type
Feature
Clarity
Needs clarification
Activity status
Stale
Tech stack
r
Domain
performance

Research direction

Start by reviewing the timing experiment at commit fbb651005e9a1b51cccbcefc29f39d90fcd376f5 and the long-vector work in issue #3957. Determine where vector copying is implemented and how integer, logical, and double vectors are covered. Done means a supported parallel copy approach is implemented and its performance and variance are measured against the listed baselines.

Written by the indexing model from the issue text.

Description

openmp performance

This might be not that high priority now, but once long vectors will be supported #3957, then having faster copy might be saving more time.
I made an experiement how much we can save on 2e9 integer vector (7.5 GB in memory).
It should be the same for logical vector, double vector (8 bytes) should get even more speed up.

  • plain for loop: 5.05s
  • memcpy: 5.3s
  • omp for static 40th one-by-one: 0.55s
  • omp for dynamic 40th memcpy 40 batches: 0.52s

memcpy timings have significantly bigger variance

code used for timings: https://gitlab.com/jangorecki/data.table/-/commit/fbb651005e9a1b51cccbcefc29f39d90fcd376f5

Dominant language
R
Stars
3.9k
Forks
1.1k
Avg merge
14h 4m
Merged PRs (30d)
4

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from Rdatatable/data.table

All issues in Rdatatable/data.table

Similar issues

More R issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.