Why is matrix multiplication faster with numpy than with ctypes in Python?

by

NumPy uses a highly-optimized, carefully-tuned BLAS method for matrix multiplication (see also: ATLAS). The specific function in this case is GEMM (for generic matrix multiplication). You can look up the original by searching for dgemm.f (it’s in Netlib).

The optimization, by the way, goes beyond compiler optimizations. Above, Philip mentioned Coppersmith–Winograd. If I remember correctly, this is the algorithm which is used for most cases of matrix multiplication in ATLAS (though a commenter notes it could be Strassen’s algorithm).

In other words, your matmult algorithm is the trivial implementation. There are faster ways to do the same thing.

More Related Contents:

  1. Why is reading lines from stdin much slower in C++ than Python?
  2. How to generate random binary numbers 0 or 1 with length of N and with option to control the probability of having 0 or 1?
  3. Why does Python code run faster in a function?
  4. Approximating data with a multi segment cubic bezier curve and a distance as well as a curvature contraint
  5. Difference between numpy dot() and Python 3.5+ matrix multiplication @
  6. how does multiplication differ for NumPy Matrix vs Array classes?
  7. Why is looping over range() in Python faster than using a while loop?
  8. Speed comparison with Project Euler: C vs Python vs Erlang vs Haskell
  9. It is more efficient to use if-return-return or if-else-return?
  10. Python (and Python C API): __new__ versus __init__
  11. How do you call Python code from C code?
  12. Extending setuptools extension to use CMake in setup.py?
  13. How do I convert a Python list into a C array by using ctypes?
  14. How do I get monotonic time durations in python?
  15. What does “Symbol not found / Expected in: flat namespace” actually mean?
  16. Why is json.loads an order of magnitude faster than ast.literal_eval?
  17. Run a python script with arguments
  18. Can we use C code in Python?
  19. How to pass a 2d array from Python to C?
  20. Python ? (conditional/ternary) operator for assignments [duplicate]
  21. Difference between np.int, np.int_, int, and np.int_t in cython?
  22. Using Cython To Link Python To A Shared Library
  23. Equivalent to GetTickCount() on Linux
  24. Calling Python script from C++ and using its output
  25. Wrapping a pre-initialized pointer in a cython class
  26. Building Mesa for windows 7. Mesa 9.1
  27. C++ conversion from NumPy array to Mat (OpenCV)
  28. ImportError: dynamic module does not define init function (initfizzbuzz)
  29. SWIG interfacing C library to Python (Creating ‘iterable’ Python data type from C ‘sequence’ struct)
  30. Embedding python in multithreaded C application

Leave a Comment