MAC M1异常缓慢

Question

我已经为我的numpy进行了简单的速度测试：

import numpy as np

A = np.random.rand(1000, 1000)
B = np.random.rand(1000, 1000)

%timeit A.dot(B)

结果是：

30.3 ms ± 829 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)

与其他人通常看到的结果相比，这个结果似乎异常缓慢（平均少于10 ms）。我想知道这种行为的原因可能是什么。

我的系统是Macos Big Sur上的M1芯片。 Python版本为3.8.13，Numpy版本为1.22.4。 Numpy是通过

pip install "numpy==1.22.4"

np.show_config（） is：

openblas64__info:
    libraries = ['openblas64_', 'openblas64_']
    library_dirs = ['/usr/local/lib']
    language = c
    define_macros = [('HAVE_CBLAS', None), ('BLAS_SYMBOL_SUFFIX', '64_'), ('HAVE_BLAS_ILP64', None)]
    runtime_library_dirs = ['/usr/local/lib']
blas_ilp64_opt_info:
    libraries = ['openblas64_', 'openblas64_']
    library_dirs = ['/usr/local/lib']
    language = c
    define_macros = [('HAVE_CBLAS', None), ('BLAS_SYMBOL_SUFFIX', '64_'), ('HAVE_BLAS_ILP64', None)]
    runtime_library_dirs = ['/usr/local/lib']
openblas64__lapack_info:
    libraries = ['openblas64_', 'openblas64_']
    library_dirs = ['/usr/local/lib']
    language = c
    define_macros = [('HAVE_CBLAS', None), ('BLAS_SYMBOL_SUFFIX', '64_'), ('HAVE_BLAS_ILP64', None), ('HAVE_LAPACKE', None)]
    runtime_library_dirs = ['/usr/local/lib']
lapack_ilp64_opt_info:
    libraries = ['openblas64_', 'openblas64_']
    library_dirs = ['/usr/local/lib']
    language = c
    define_macros = [('HAVE_CBLAS', None), ('BLAS_SYMBOL_SUFFIX', '64_'), ('HAVE_BLAS_ILP64', None), ('HAVE_LAPACKE', None)]
    runtime_library_dirs = ['/usr/local/lib']
Supported SIMD extensions in this NumPy install:
    baseline = SSE,SSE2,SSE3
    found = SSSE3,SSE41,POPCNT,SSE42
    not found = AVX,F16C,FMA3,AVX2,AVX512F,AVX512CD,AVX512_KNL,AVX512_SKX,AVX512_CLX,AVX512_CNL,AVX512_ICL

---

edit：

我对此代码shippet进行了另一项测试（来自 1 ）：

import time
import numpy as np
np.random.seed(42)
a = np.random.uniform(size=(300, 300))
runtimes = 10

timecosts = []
for _ in range(runtimes):
    s_time = time.time()
    for i in range(100):
        a += 1
        np.linalg.svd(a)
    timecosts.append(time.time() - s_time)

print(f'mean of {runtimes} runs: {np.mean(timecosts):.5f}s')

我的测试结果是：

mean of 10 runs: 6.17438s

而网站上的参考结果 1 是：（芯片是m1 max），

+-----------------------------------+-----------------------+--------------------+
|   Python installed by (run on)→   | Miniforge (native M1) | Anaconda (Rosseta) |
+----------------------+------------+------------+----------+----------+---------+
| Numpy installed by ↓ | Run from → |  Terminal  |  PyCharm | Terminal | PyCharm |
+----------------------+------------+------------+----------+----------+---------+
|          Apple Tensorflow         |   4.19151  |  4.86248 |     /    |    /    |
+-----------------------------------+------------+----------+----------+---------+
|        conda install numpy        |   4.29386  |  4.98370 |  4.10029 | 4.99271 |
+-----------------------------------+------------+----------+----------+---------+

从结果，我的时间，我的时间安排与参考中的任何numpy版本相比，代码较慢。

Answer 1

我注意到M1上的类似放缓，但是我认为至少在我的计算机上的实际原因并不是从根本上有故障的Numpy安装，而是基准本身的一些问题。考虑以下示例：

In [25]: from scipy import linalg

In [26]: a = np.random.randn(1000,100)

In [27]: %timeit a.T @ a
226 µs ± 7.03 µs per loop (mean ± std. dev. of 7 runs, 1,000 loops each)

In [28]: x = a.T @ a

In [29]: %timeit linalg.eigh(x)
1.69 ms ± 88.8 µs per loop (mean ± std. dev. of 7 runs, 1,000 loops each)

In [30]: %timeit linalg.eigh(a.T @ a)
428 ms ± 99.9 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

computing x = at @ a; eigh（x）需要2 ms，而eigh（在 @ a） 400 ms。我认为在后一种情况下，％timeit是一些问题。也许由于某种原因，计算被路由到“效率核心”？

我的初步答案是，您的第一个基准标准>％TimeIt是不可靠的。

Answer 2

如果您怀疑在TimeIt中存在问题，请尝试使用时间代替

import time
start = time.time()

# your numpy test here

took=time.time() - start
print("Test took "+str(took)+" seconds.")

Apple Silicon上的Numpy的更多信息，请阅读链接波纹管中的第一个答案。为了获得最佳性能，建议使用Apple的加速Veclib。如果您使用conda安装，请访问 @andrejhribernik的评论：
为什么M1 Max上的本地人比Old Intel i5上的Python慢​​得多？