Coder Social home page Coder Social logo

Comments (1)

giaf avatar giaf commented on July 22, 2024 1

In general the BLASFEO own API is closely related to the standard BLAS API, with the key difference that there is an additional argument for the result, such that the API is "non destructive".
The standard BLAS API is "destructive" and an explicit matrix copy has to be performed in the rather common case the input/output argument should be preserved.
So the BLASFEO API gives a bit more flexibility, and it is always possible to have the added output argument coincide with the argument acting as the input/output in the standard BLAS API.

E.g. if you take the dgemm in the standard BLAS API, it is something like
C = beta * C + alpha * A * B
then in BLASFEO there would be the additional argument D for the result
D = beta * C + alpha * A * B
that can possibly (but not necessarily) coincide with C.

Similarly, for dtrsm the standard BLAS API would be
B = alpha * inv(A) * B
and in the BLASFEO API again there is an additional argument D for the result
D = alpha * inv(A) * B
that can possibly (but not necessarily) coincide with B.

All other limitations about arguments overlapping are analogue to the standard BLAS API, where BTW are also not so clearly described but rather implicit in the understanding of the operation performed.
E.g. in case A and C coincide in dgemm in the BLAS API, clearly A would be overwritten with the result as soon as this is computed a bit at a time, so the result would be wrong in an unpredictable and implementation-dependent way.

from blasfeo.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.