Describe the issue:

I am testing this code:

import numpy as np
a = np.random.rand(1000, 1000)
ainv = np.linalg.inv(a)

on an Intel ultra 7 processor (155H).
This should take approx 30-50ms, but it takes 12000ms on numpy 1.26.4 or 2.0 (using pip install numpy). This is really crazy slow!!!

If I unstalled the version 1.26.3 with MKL support from Gohlke:
https://github.com/cgohlke/numpy-mkl-wheels/releases/tag/v2024.1.3
the timing was appox 2000ms, whic is still 100x more than normal.

If I installed Anaconda (which I do not prefer), I have reached 30-50ms.

I am doing anything wrong?

Reproduce the code example:

import numpy as np
a = np.random.rand(1000, 1000)
ainv = np.linalg.inv(a)

Error message:

No response

Python and NumPy Versions:

1.26.3
3.12.5 (tags/v3.12.5:ff3bc82, Aug 6 2024, 20:45:27) [MSC v.1940 64 bit (AMD64)]

Runtime Environment:

[{'numpy_version': '1.26.4',
'python': '3.12.5 (tags/v3.12.5:ff3bc82, Aug 6 2024, 20:45:27) [MSC v.1940 '
'64 bit (AMD64)]',
'uname': uname_result(system='Windows', node='Spectre-JS', release='11', version='10.0.22631', machine='AMD64')},
{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'],
'found': ['SSSE3',
'SSE41',
'POPCNT',
'SSE42',
'AVX',
'F16C',
'FMA3',
'AVX2'],
'not_found': ['AVX512F',
'AVX512CD',
'AVX512_SKX',
'AVX512_CLX',
'AVX512_CNL',
'AVX512_ICL']}},
{'architecture': 'Haswell',
'filepath': 'C:\Users\janko\AppData\Local\Programs\Python\Python312\Lib\site-packages\numpy.libs\libopenblas64__v0.3.23-293-gc2f4bdbb-gcc_10_3_0-2bde3a66a51006b2b53eb373ff767a3f.dll',
'internal_api': 'openblas',
'num_threads': 22,
'prefix': 'libopenblas',
'threading_layer': 'pthreads',
'user_api': 'blas',
'version': '0.3.23.dev'}]

Context for the issue:

Matrix inversion is essential to me. Such slow inversion makes my code unusable or I have to go for Anaconda.