Skip to content

Fully mutating linear solver #48

Open
Open
Fully mutating linear solver#48
Labels
performanceImprove performance or stability

Description

@gdalle
gdalle
opened on May 24, 2023

At the moment, AbstractDifferentiation does not have a fully mutating API, and benchmarks show allocations are non-negligible when multiplying by A and B. Opening this to track JuliaDiff/AbstractDifferentiation.jl#14 and see what we can do

Activity

  1. gdalle
    added
    featureNew feature or request
    on May 24, 2023

  2. mohdibntarek commented on May 24, 2023

    @mohdibntarek
    mohdibntarek
    on May 24, 2023
    Collaborator

    The problem is that most AD packages don't actually support mutation.

  3. gdalle commented on May 24, 2023

    @gdalle
    gdalle
    on May 24, 2023
    MemberAuthor

    Yeah I know, it would be useless for ChainRules, but not for ForwardDiff?

  4. gdalle
    added
    performanceImprove performance or stability
    and removed
    featureNew feature or request
    on Aug 8, 2023

  5. gdalle commented on Feb 21, 2024

    @gdalle
    gdalle
    on Feb 21, 2024
    MemberAuthor

    Trying to solve this with https://github.com/gdalle/DifferentiationInterface.jl

  6. gdalle commented on Apr 10, 2024

    @gdalle
    gdalle
    on Apr 10, 2024
    MemberAuthor

    This is now possible with DifferentiationInterface but

    • it won't work with every outer backend
    • it requires a mutating version of the conditions conditions!(c, x, y)

    I hope to make it part of v0.7

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    performanceImprove performance or stability

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions