R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out" Copyright (C) 2020 The R Foundation for Statistical Computing Platform: arm-apple-darwin20.2.0 (64-bit)
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests): 3,500,000 Fibonacci numbers calculation (vector calc): 0.112 (sec). Grand common divisors of 1,000,000 pairs (recursion): 0.26 (sec). Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.146 (sec). Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 0.597 (sec). Escoufier's method on a 60 x 60 matrix (mixed): 0.598 (sec). # Matrix calculation benchmarks (5 tests): Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.245 (sec). 2,500 x 2,500 normal distributed random matrix^1,000: 0.113 (sec). Sorting of 7,000,000 random values: 0.606 (sec). 2,500 x 2,500 cross-product matrix (b = a' * a): 9.54 (sec). Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.799 (sec). # Matrix function benchmarks (5 tests): Cholesky decomposition of a 3,000 x 3,000 matrix: 5.21 (sec). Determinant of a 2,500 x 2,500 random matrix: 1.78 (sec). Eigenvalues of a 640 x 640 random matrix: 0.476 (sec). FFT over 2,500,000 random values: 0.113 (sec). Inverse of a 1,600 x 1,600 random matrix: 1.45 (sec).
R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out" Copyright (C) 2020 The R Foundation for Statistical Computing Platform: x86_64-pc-linux-gnu (64-bit)
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests): 3,500,000 Fibonacci numbers calculation (vector calc): 0.221 (sec). Grand common divisors of 1,000,000 pairs (recursion): 0.537 (sec). Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.285 (sec). Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.29 (sec). Escoufier's method on a 60 x 60 matrix (mixed): 1.25 (sec). # Matrix calculation benchmarks (5 tests): Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.726 (sec). 2,500 x 2,500 normal distributed random matrix^1,000: 0.287 (sec). Sorting of 7,000,000 random values: 0.767 (sec). 2,500 x 2,500 cross-product matrix (b = a' * a): 0.241 (sec). Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.0273 (sec). # Matrix function benchmarks (5 tests): Cholesky decomposition of a 3,000 x 3,000 matrix: 0.225 (sec). Determinant of a 2,500 x 2,500 random matrix: 0.187 (sec). Eigenvalues of a 640 x 640 random matrix: 0.42 (sec). FFT over 2,500,000 random values: 0.396 (sec). Inverse of a 1,600 x 1,600 random matrix: 0.186 (sec).
这种小测试应该和内存没关系，主要是单核性能i5 6400+32GB内存,windows不用 openblas之类的多线程blas： > res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests): 3,500,000 Fibonacci numbers calculation (vector calc): 0.767 (sec). Grand common divisors of 1,000,000 pairs (recursion): 0.83 (sec). Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.313 (sec). Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.34 (sec). Escoufier's method on a 60 x 60 matrix (mixed): 1.02 (sec). # Matrix calculation benchmarks (5 tests): Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.757 (sec). 2,500 x 2,500 normal distributed random matrix^1,000: 0.917 (sec). Sorting of 7,000,000 random values: 0.823 (sec). 2,500 x 2,500 cross-product matrix (b = a' * a): 11.4 (sec). Linear regr. over a 5,000 x 500 matrix (c = a \ b'): 0.9 (sec). # Matrix function benchmarks (5 tests): Cholesky decomposition of a 3,000 x 3,000 matrix: 6.19 (sec). Determinant of a 2,500 x 2,500 random matrix: 2.21 (sec). Eigenvalues of a 640 x 640 random matrix: 0.613 (sec). FFT over 2,500,000 random values: 0.327 (sec). Inverse of a 1,600 x 1,600 random matrix: 2.18 (sec).
loaded via a namespace (and not attached): [1] compiler_4.0.3 > res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests): 3,500,000 Fibonacci numbers calculation (vector calc): 0.126 (sec). Grand common divisors of 1,000,000 pairs (recursion): 0.418 (sec). Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.169 (sec). Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.16 (sec). Escoufier's method on a 60 x 60 matrix (mixed): 0.515 (sec). # Matrix calculation benchmarks (5 tests): Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.432 (sec). 2,500 x 2,500 normal distributed random matrix^1,000: 0.152 (sec). Sorting of 7,000,000 random values: 0.496 (sec). 2,500 x 2,500 cross-product matrix (b = a' * a): 0.0723 (sec). Linear regr. over a 5,000 x 500 matrix (c = a \ b'): 0.008 (sec). # Matrix function benchmarks (5 tests): Cholesky decomposition of a 3,000 x 3,000 matrix: 0.095 (sec). Determinant of a 2,500 x 2,500 random matrix: 0.07 (sec). Eigenvalues of a 640 x 640 random matrix: 0.205 (sec). FFT over 2,500,000 random values: 0.154 (sec). Inverse of a 1,600 x 1,600 random matrix: 0.0723 (sec)
【 在 lix0519 (tgh1122334) 的大作中提到: 】 : 这种小测试应该和内存没关系，主要是单核性能i5 6400+32GB内存,windows不用 : openblas之类的多线程blas： : > res <- benchmarkme::benchmark_std="">: # Programming benchmarks (5 tests): : 3,500,000 Fibonacci numbers calculation (vector calc): 0.767 (sec). : Grand common divisors of 1,000,000 pairs (recursion): 0.83 (sec). : Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.313 (sec). : Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.34 (sec). : Escoufier's method on a 60 x 60 matrix (mixed): 1.02 (sec). : # Matrix calculation benchmarks (5 tests): : ...................
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests): 3,500,000 Fibonacci numbers calculation (vector calc): 0.148 (sec). Grand common divisors of 1,000,000 pairs (recursion): 0.328 (sec). Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.154 (sec). Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 0.861 (sec). Escoufier's method on a 60 x 60 matrix (mixed): 0.553 (sec). # Matrix calculation benchmarks (5 tests): Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.361 (sec). 2,500 x 2,500 normal distributed random matrix^1,000: 0.149 (sec). Sorting of 7,000,000 random values: 0.491 (sec). 2,500 x 2,500 cross-product matrix (b = a' * a): 0.0617 (sec). Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.0103 (sec). # Matrix function benchmarks (5 tests): Cholesky decomposition of a 3,000 x 3,000 matrix: 0.074 (sec). Determinant of a 2,500 x 2,500 random matrix: 0.0607 (sec). Eigenvalues of a 640 x 640 random matrix: 0.165 (sec). FFT over 2,500,000 random values: 0.154 (sec). Inverse of a 1,600 x 1,600 random matrix: 0.055 (sec).
这个果粉在https://forums.macrumors.com/threads/data-science-r-and-spss-26-etc-under-
rosetta-2-apple-silicon-m1.2269302/?post=29326680#post-29326680
发表了16GB Mac Pro上的R benchmark
R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
Copyright (C) 2020 The R Foundation for Statistical Computing
Platform: arm-apple-darwin20.2.0 (64-bit)
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests):
3,500,000 Fibonacci numbers calculation (vector calc): 0.112 (sec).
Grand common divisors of 1,000,000 pairs (recursion): 0.26 (sec).
Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.146 (sec).
Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 0.597 (sec).
Escoufier's method on a 60 x 60 matrix (mixed): 0.598 (sec).
# Matrix calculation benchmarks (5 tests):
Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.245 (sec).
2,500 x 2,500 normal distributed random matrix^1,000: 0.113 (sec).
Sorting of 7,000,000 random values: 0.606 (sec).
2,500 x 2,500 cross-product matrix (b = a' * a): 9.54 (sec).
Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.799 (sec).
# Matrix function benchmarks (5 tests):
Cholesky decomposition of a 3,000 x 3,000 matrix: 5.21 (sec).
Determinant of a 2,500 x 2,500 random matrix: 1.78 (sec).
Eigenvalues of a 640 x 640 random matrix: 0.476 (sec).
FFT over 2,500,000 random values: 0.113 (sec).
Inverse of a 1,600 x 1,600 random matrix: 1.45 (sec).
看了一下，肯定是没法吊打i9移动的
因为我几年前的挤牙膏的8GB i5上网本（无独立GPU)分数如下
一些项目慢一倍可以理解，毕竟几年前的CPU，核数比m1少一半, 内存少一半
但是要快，就可能快几十倍
难怪果粉死活不热烈张贴常见benchmark分数
m1估计至少要到m2才能成熟
R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
Copyright (C) 2020 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests):
3,500,000 Fibonacci numbers calculation (vector calc): 0.221 (sec).
Grand common divisors of 1,000,000 pairs (recursion): 0.537 (sec).
Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.285 (sec).
Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.29 (sec).
Escoufier's method on a 60 x 60 matrix (mixed): 1.25 (sec).
# Matrix calculation benchmarks (5 tests):
Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.726 (sec).
2,500 x 2,500 normal distributed random matrix^1,000: 0.287 (sec).
Sorting of 7,000,000 random values: 0.767 (sec).
2,500 x 2,500 cross-product matrix (b = a' * a): 0.241 (sec).
Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.0273 (sec).
# Matrix function benchmarks (5 tests):
Cholesky decomposition of a 3,000 x 3,000 matrix: 0.225 (sec).
Determinant of a 2,500 x 2,500 random matrix: 0.187 (sec).
Eigenvalues of a 640 x 640 random matrix: 0.42 (sec).
FFT over 2,500,000 random values: 0.396 (sec).
Inverse of a 1,600 x 1,600 random matrix: 0.186 (sec).
M1肯定不行的，看看M1X吧
这种小测试应该和内存没关系，主要是单核性能i5 6400+32GB内存,windows不用
openblas之类的多线程blas：
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests):
3,500,000 Fibonacci numbers calculation (vector calc): 0.767 (sec).
Grand common divisors of 1,000,000 pairs (recursion): 0.83 (sec).
Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.313 (sec).
Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.34 (sec).
Escoufier's method on a 60 x 60 matrix (mixed): 1.02 (sec).
# Matrix calculation benchmarks (5 tests):
Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.757 (sec).
2,500 x 2,500 normal distributed random matrix^1,000: 0.917 (sec).
Sorting of 7,000,000 random values: 0.823 (sec).
2,500 x 2,500 cross-product matrix (b = a' * a): 11.4 (sec).
Linear regr. over a 5,000 x 500 matrix (c = a \ b'): 0.9 (sec).
# Matrix function benchmarks (5 tests):
Cholesky decomposition of a 3,000 x 3,000 matrix: 6.19 (sec).
Determinant of a 2,500 x 2,500 random matrix: 2.21 (sec).
Eigenvalues of a 640 x 640 random matrix: 0.613 (sec).
FFT over 2,500,000 random values: 0.327 (sec).
Inverse of a 1,600 x 1,600 random matrix: 2.18 (sec).
Windows可以用openblas或者mkl
WSL2或者docker
最简单地，docker pull rocker/rstudio 这是openblas的
或者不想玩docker wsl, 直接下载一个microsoft open R 这是mkl的，但是优化不好
还不如上边
那个块
单核比多核，这个测试能差好几倍
以下是桌面i7 9700k 64GB openblas的benchmark
Matrix products: default
BLAS/LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.8.so
loaded via a namespace (and not attached):
[1] compiler_4.0.3
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests):
3,500,000 Fibonacci numbers calculation (vector calc): 0.126 (sec).
Grand common divisors of 1,000,000 pairs (recursion): 0.418 (sec).
Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.169 (sec).
Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.16 (sec).
Escoufier's method on a 60 x 60 matrix (mixed): 0.515 (sec).
# Matrix calculation benchmarks (5 tests):
Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.432 (sec).
2,500 x 2,500 normal distributed random matrix^1,000: 0.152 (sec).
Sorting of 7,000,000 random values: 0.496 (sec).
2,500 x 2,500 cross-product matrix (b = a' * a): 0.0723 (sec).
Linear regr. over a 5,000 x 500 matrix (c = a \ b'): 0.008 (sec).
# Matrix function benchmarks (5 tests):
Cholesky decomposition of a 3,000 x 3,000 matrix: 0.095 (sec).
Determinant of a 2,500 x 2,500 random matrix: 0.07 (sec).
Eigenvalues of a 640 x 640 random matrix: 0.205 (sec).
FFT over 2,500,000 random values: 0.154 (sec).
Inverse of a 1,600 x 1,600 random matrix: 0.0723 (sec)
【 在 lix0519 (tgh1122334) 的大作中提到: 】
: 这种小测试应该和内存没关系，主要是单核性能i5 6400+32GB内存,windows不用
: openblas之类的多线程blas：
: > res <- benchmarkme::benchmark_std="">: # Programming benchmarks (5 tests):
: 3,500,000 Fibonacci numbers calculation (vector calc): 0.767 (sec).
: Grand common divisors of 1,000,000 pairs (recursion): 0.83 (sec).
: Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.313 (sec).
: Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 1.34 (sec).
: Escoufier's method on a 60 x 60 matrix (mixed): 1.02 (sec).
: # Matrix calculation benchmarks (5 tests):
: ...................
mkl比openblas稍好：
> res <- benchmarkme::benchmark_std=""># Programming benchmarks (5 tests):
3,500,000 Fibonacci numbers calculation (vector calc): 0.148 (sec).
Grand common divisors of 1,000,000 pairs (recursion): 0.328 (sec).
Creation of a 3,500 x 3,500 Hilbert matrix (matrix calc): 0.154 (sec).
Creation of a 3,000 x 3,000 Toeplitz matrix (loops): 0.861 (sec).
Escoufier's method on a 60 x 60 matrix (mixed): 0.553 (sec).
# Matrix calculation benchmarks (5 tests):
Creation, transp., deformation of a 5,000 x 5,000 matrix: 0.361 (sec).
2,500 x 2,500 normal distributed random matrix^1,000: 0.149 (sec).
Sorting of 7,000,000 random values: 0.491 (sec).
2,500 x 2,500 cross-product matrix (b = a' * a): 0.0617 (sec).
Linear regr. over a 5,000 x 500 matrix (c = a b'): 0.0103 (sec).
# Matrix function benchmarks (5 tests):
Cholesky decomposition of a 3,000 x 3,000 matrix: 0.074 (sec).
Determinant of a 2,500 x 2,500 random matrix: 0.0607 (sec).
Eigenvalues of a 640 x 640 random matrix: 0.165 (sec).
FFT over 2,500,000 random values: 0.154 (sec).
Inverse of a 1,600 x 1,600 random matrix: 0.055 (sec).
【 在 didadida (滴滴嗒嗒) 的大作中提到: 】
: Windows可以用openblas或者mkl
: WSL2或者docker
: 最简单地，docker pull rocker/rstudio
: 单核比多核，这个测试能差好几倍(最下边的Windows Default和Ubuntu Default都是单
: 核）
: R-benchmark-25.R Total time (sec)
: Windows Intel MKL 6.95
: Ubuntu Intel MKL 7.06
: Ubuntu OpenBLAS 7.41
: Ubuntu ATLAS 10.37
: ...................
至于么，我都替你累得慌
哎呦卧槽，一股农民味儿扑面而来。
Mac Pro
MacBook Pro
滴弟弟傻傻分不清楚
无所谓啊，知道评测的是Pro不是air或者mini就行了
只有傻傻的果粉才会企图区分“科学计算“和”数学计算“，却又连mkl和matlab都区
分不了
这才是真傻
【 在 ToyotaYaris (丰田雅力士) 的大作中提到: 】
: 哎呦卧槽，一股农民味儿扑面而来。
: Mac Pro
: MacBook Pro
: 滴弟弟傻傻分不清楚
统计傻屌算文科
【 在 didadida (滴滴嗒嗒) 的大作中提到: 】
: 无所谓啊，知道评测的是Pro不是air或者mini就行了
: 只有傻傻的果粉才会企图区分“科学计算“和”数学计算“，却又连mkl和matlab都区
: 分不了
: 这才是真傻
你一个连matlab和mkl都分不清楚的文科生就不用笑话统计是文科了
当然，你可以吹嘘自己是杨振宁，是张益唐
因为互联网上的老话是“on the internet nobody knows you are a dog"
【 在 ToyotaYaris (丰田雅力士) 的大作中提到: 】
: 统计傻屌算文科
统计真的是文科
【 在 didadida (滴滴嗒嗒) 的大作中提到: 】
: 你一个连matlab和mkl都分不清楚的文科生就不用笑话统计是文科了
: 当然，你可以吹嘘自己是杨振宁，是张益唐
: 因为互联网上的老话是“on the internet nobody knows you are a dog"
统计是文科，那又怎么着？
【 在 ToyotaYaris (丰田雅力士) 的大作中提到: 】
: 统计真的是文科
果粉sb鉴定完毕？
这个不是 M1 性能差,是 openblas 还没移植到 M1 上吧,所以后两项差.
【 在 didadida (滴滴嗒嗒) 的大作中提到: 】
: 这个果粉在
: https://forums.macrumors.com/threads/data-science-r-and-spss-26-etc-under-: rosetta-2-apple-silicon-m1.2269302/?post=29326680#post-29326680
: 发表了16GB Mac Pro上的R benchmark
: R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
: Copyright (C) 2020 The R Foundation for Statistical Computing
: Platform: arm-apple-darwin20.2.0 (64-bit)
: > res <- benchmarkme::benchmark_std="">: # Programming benchmarks (5 tests):
: 3,500,000 Fibonacci numbers calculation (vector calc): 0.112 (sec).
: ...................