Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Just for normies like me who don’t know what Simon meant :-) 1. Open terminal.app and cd to where the relevant files are to keep commands manageable in length cd /Library/Frameworks/R.framework/Resources/lib/ 2. Find the paths to different versions of veclib by ls’ing the directory they should live in: ls -l libRblas*dylib The existing symbolic link will look like this lrwxr-xr-x 1 root admin 16 24 Sep 12:10 libRblas.dylib -> libRblas.0.dylib Which means calls to “libRblas.dylib” will resolve to "libRblas.0.dylib" 3. Overwrite this link with a new one from Apple's veclib (libRblas.vecLib.dylib ) to libRblas.dylib ln -s -i -v libRblas.vecLib.dylib libRblas.dylib When you restart R, it should now call libRblas.vecLib.dylib for math To switch back cd /Library/Frameworks/R.framework/Resources/lib/ ln -s -i -v libRblas.0.dylib libRblas.dylib t PS: The performance boost is staggering!! I. Matrix calculationApple Blas (Sec)Default Blas (Sec) gain (loss) Creation, transp., deformation of a 2500x2500 matrix0.546 0.672 123% 2400x2400 normal distributed random matrix ^10000.139 0.139 100% Sorting of 7,000,000 random values 0.600 0.605 101% 2800x2800 cross-product matrix (b = a' * a) 0.210 9.061 4315% !! Linear regr. over a 3000x3000 matrix (c = a \ b') 0.134 4.379 3276% Trimmed geom. mean (2 extremes eliminated) 0.252 1.212 482% II. Matrix functions FFT over 2,400,000 random values0.198 0.183 92% Eigenvalues of a 640x640 random matrix 0.295 0.625 212% Determinant of a 2500x2500 random matrix0.146 2.891 1980% !! Cholesky decomposition of a 3000x3000 matrix0.222 3.772 1696% !! Inverse of a 1600x1600 random matrix0.284 2.417 852% Trimmed geom. mean (2 extremes eliminated) 0.232 1.634 704% III. Programmation 3,500,000 Fibonacci numbers calculation (vector calc) 0.204 0.204 100% Creation of a 3000x3000 Hilbert matrix (matrix calc)0.232 0.211 91% Grand common divisors of 400,000 pairs (recursion) 0.274 0.243 89% Creation of a 500x500 Toeplitz matrix (loops) 0.038 0.041 106% Escoufier's method on a 45x45 matrix (mixed)0.260 0.307 118% Trimmed geom. mean (2 extremes eliminated) 0.231 0.219 95% Total time for all 15 tests 3.782 25.749 681% Overall mean (sum of I, II and III trimmed means/3) 0.238 0.757 318% The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. Is e buidheann carthannais a th’ ann an Oilthigh Dhùn Èideann, clàraichte an Alba, àireamh clàraidh SC005336. [[alternative HTML version deleted]] ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Out of curiosity, do we know if vecLib/Accelerate has been optimized for M1? On Sun, Oct 31, 2021 at 10:00 PM Kieran Healy wrote: > Thanks, Simon. > > With the vecLib/Accelerate BLAS, the results are indeed rather faster :) > > Kieran > > > 14” MacBook Pro / M1 Max > >R Benchmark 2.5 >=== > Number of times each test is run__: 3 > >I. Matrix calculation >- > Creation, transp., deformation of a 2500x2500 matrix (sec): > 0.260 > 2400x2400 normal distributed random matrix ^1000 (sec): 0.105 > Sorting of 7,000,000 random values__ (sec): 0.595 > 2800x2800 cross-product matrix (b = a' * a)_ (sec): > 0.056 > Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): > 0.0448 > > Trimmed geom. mean (2 extremes eliminated): > 0.115802825957684 > >II. Matrix functions > > FFT over 2,400,000 random values (sec): > 0.07233329 > Eigenvalues of a 640x640 random matrix__ (sec): > 0.157 > Determinant of a 2500x2500 random matrix (sec): > 0.098 > Cholesky decomposition of a 3000x3000 matrix (sec): > 0.07166654 > Inverse of a 1600x1600 random matrix (sec): > 0.0826667 > > Trimmed geom. mean (2 extremes eliminated): > 0.0839655943753058 > >III. Programmation >-- > 3,500,000 Fibonacci numbers calculation (vector calc)(sec): > 0.0937 > Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): > 0.1123332 > Grand common divisors of 400,000 pairs (recursion)__ (sec): > 0.07766657 > Creation of a 500x500 Toeplitz matrix (loops)___ (sec): > 0.0172 > Escoufier's method on a 45x45 matrix (mixed) (sec): > 0.1119998 > > Trimmed geom. mean (2 extremes eliminated): > 0.0932888677080541 > > > Total time for all 15 tests_ (sec): > 1.957333 > Overall mean (sum of I, II and III trimmed means/3)_ (sec): > 0.0968018035139188 > --- End of test --- > > > On Oct 31, 2021, at 9:03 PM, Simon Urbanek > wrote: > > > > Kieran, > > > > the reference benchmarks have been calibrated against vecLib/Accelerate > BLAS. If you use reference BLAS it can be a lot slower. You can switch > between reference BLAS and vecLib in R CRAN releases simply by switching > the libRblas.dylib symlink (in $R_HOME/lib), e.g.: > > > > ls -l /Library/Frameworks/R.framework/Resources/lib/libRblas*dylib > > -rwxrwxr-x 1 root admin 226288 Oct 31 14:41 > /Library/Frameworks/R.framework/Resources/lib/libRblas.0.dylib > > lrwxr-xr-x 1 root.admin 21 Nov 1 09:56 > /Library/Frameworks/R.framework/Resources/lib/libRblas.dylib -> > libRblas.vecLib.dylib > > -rwxrwxr-x 1 root admin 154368 Oct 31 14:41 > /Library/Frameworks/R.framework/Resources/lib/libRblas.vecLib.dylib > > > > (For recent R you'll need R 4.1.1 or higher) > > > > Cheers, > > Simon > > > > PS: reminder to everyone, please test R 4.1.2 RC - now are the last few > hours to report anything! > > > > > > ___ > R-SIG-Mac mailing list > R-SIG-Mac@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-mac > -- Best, Kasper [[alternative HTML version deleted]] ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Thanks, Simon. With the vecLib/Accelerate BLAS, the results are indeed rather faster :) Kieran 14” MacBook Pro / M1 Max R Benchmark 2.5 === Number of times each test is run__: 3 I. Matrix calculation - Creation, transp., deformation of a 2500x2500 matrix (sec): 0.260 2400x2400 normal distributed random matrix ^1000 (sec): 0.105 Sorting of 7,000,000 random values__ (sec): 0.595 2800x2800 cross-product matrix (b = a' * a)_ (sec): 0.056 Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): 0.0448 Trimmed geom. mean (2 extremes eliminated): 0.115802825957684 II. Matrix functions FFT over 2,400,000 random values (sec): 0.07233329 Eigenvalues of a 640x640 random matrix__ (sec): 0.157 Determinant of a 2500x2500 random matrix (sec): 0.098 Cholesky decomposition of a 3000x3000 matrix (sec): 0.07166654 Inverse of a 1600x1600 random matrix (sec): 0.0826667 Trimmed geom. mean (2 extremes eliminated): 0.0839655943753058 III. Programmation -- 3,500,000 Fibonacci numbers calculation (vector calc)(sec): 0.0937 Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): 0.1123332 Grand common divisors of 400,000 pairs (recursion)__ (sec): 0.07766657 Creation of a 500x500 Toeplitz matrix (loops)___ (sec): 0.0172 Escoufier's method on a 45x45 matrix (mixed) (sec): 0.1119998 Trimmed geom. mean (2 extremes eliminated): 0.0932888677080541 Total time for all 15 tests_ (sec): 1.957333 Overall mean (sum of I, II and III trimmed means/3)_ (sec): 0.0968018035139188 --- End of test --- > On Oct 31, 2021, at 9:03 PM, Simon Urbanek > wrote: > > Kieran, > > the reference benchmarks have been calibrated against vecLib/Accelerate BLAS. > If you use reference BLAS it can be a lot slower. You can switch between > reference BLAS and vecLib in R CRAN releases simply by switching the > libRblas.dylib symlink (in $R_HOME/lib), e.g.: > > ls -l /Library/Frameworks/R.framework/Resources/lib/libRblas*dylib > -rwxrwxr-x 1 root admin 226288 Oct 31 14:41 > /Library/Frameworks/R.framework/Resources/lib/libRblas.0.dylib > lrwxr-xr-x 1 root.admin 21 Nov 1 09:56 > /Library/Frameworks/R.framework/Resources/lib/libRblas.dylib -> > libRblas.vecLib.dylib > -rwxrwxr-x 1 root admin 154368 Oct 31 14:41 > /Library/Frameworks/R.framework/Resources/lib/libRblas.vecLib.dylib > > (For recent R you'll need R 4.1.1 or higher) > > Cheers, > Simon > > PS: reminder to everyone, please test R 4.1.2 RC - now are the last few hours > to report anything! > > ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Kieran,
the reference benchmarks have been calibrated against vecLib/Accelerate BLAS.
If you use reference BLAS it can be a lot slower. You can switch between
reference BLAS and vecLib in R CRAN releases simply by switching the
libRblas.dylib symlink (in $R_HOME/lib), e.g.:
ls -l /Library/Frameworks/R.framework/Resources/lib/libRblas*dylib
-rwxrwxr-x 1 root admin 226288 Oct 31 14:41
/Library/Frameworks/R.framework/Resources/lib/libRblas.0.dylib
lrwxr-xr-x 1 root.admin 21 Nov 1 09:56
/Library/Frameworks/R.framework/Resources/lib/libRblas.dylib ->
libRblas.vecLib.dylib
-rwxrwxr-x 1 root admin 154368 Oct 31 14:41
/Library/Frameworks/R.framework/Resources/lib/libRblas.vecLib.dylib
(For recent R you'll need R 4.1.1 or higher)
Cheers,
Simon
PS: reminder to everyone, please test R 4.1.2 RC - now are the last few hours
to report anything!
> On Nov 1, 2021, at 11:31 AM, Kieran Healy wrote:
>
> Hello,
>
> Just out of interest, I ran benchmark-25.R from Simon’s repo, as I have
> access to an M1 Max. Are the *very* long times on cross-product, linear
> regression, and Matrix functions a consequence of the BLAS version?
>
> Kieran
>
>> sessionInfo()
> R version 4.1.1 (2021-08-10)
> Platform: aarch64-apple-darwin20 (64-bit)
> Running under: macOS Monterey 12.0.1
>
> Matrix products: default
> BLAS:
> /Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRblas.0.dylib
> LAPACK:
> /Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRlapack.dylib
>
>
>> source("R-benchmark-25.R")
> Loading required package: SuppDists
>
>
> R Benchmark 2.5
> ===
> Number of times each test is run__: 3
>
> I. Matrix calculation
> -
> Creation, transp., deformation of a 2500x2500 matrix (sec): 0.2496656
> 2400x2400 normal distributed random matrix ^1000 (sec): 0.1050009
> Sorting of 7,000,000 random values__ (sec): 0.5946673
> 2800x2800 cross-product matrix (b = a' * a)_ (sec): 13.30167
> Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): 6.270334
>
> Trimmed geom. mean (2 extremes eliminated): 0.976431082297569
>
> II. Matrix functions
>
> FFT over 2,400,000 random values (sec):
> 0.07266631
> Eigenvalues of a 640x640 random matrix__ (sec): 0.4256672
> Determinant of a 2500x2500 random matrix (sec): 1.738333
> Cholesky decomposition of a 3000x3000 matrix (sec): 5.17
> Inverse of a 1600x1600 random matrix (sec): 1.430996
>
>Trimmed geom. mean (2 extremes eliminated): 1.01925013610031
>
> III. Programmation
> --
> 3,500,000 Fibonacci numbers calculation (vector calc)(sec):
> 0.09500273
> Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): 0.1153335
> Grand common divisors of 400,000 pairs (recursion)__ (sec):
> 0.07999841
> Creation of a 500x500 Toeplitz matrix (loops)___ (sec):
> 0.01733593
> Escoufier's method on a 45x45 matrix (mixed) (sec): 0.1529963
>
>Trimmed geom. mean (2 extremes eliminated): 0.0957023962714685
>
>
> Total time for all 15 tests_ (sec): 29.823
> Overall mean (sum of I, II and III trimmed means/3)_ (sec): 0.45668322781674
> --- End of test ---
>
> Warning messages:
> 1: In remove("a", "b") : object 'a' not found
> 2: In remove("a", "b") : object 'b' not found
>
>> On Oct 31, 2021, at 5:11 PM, Simon Urbanek
>> wrote:
>>
>>
>> Tim,
>>
>> that is a great idea, those test are really old. Just for the fun of it I
>> have run the tests on my old iMac, but with R 4.1.2 and they still work.
>> It's nice to see the huge speed improvements in loops and similar (see below
>> - recall the original tests were scaled to be around 1).
>>
>> I have added the page to the repo
>> https://github.com/R-macos/R-mac-dev
>> so I'd be happy to review PRs, but I'll probably want to re-do it first so
>> it is better organized for comparisons as we have to also accommodate M1 etc.
>>
>> Cheers,
>> Simon
>>
>> ---
>> iMac14,2 3.2Ghz i5, macOS 10.4.6, R 4.1.2 vecib/Accelerate BLAS
>>
>>
>> R Benchmark 2.5
>> ===
>> Number of times each test is run__: 3
>>
>> I. Matrix calculation
>> -
>> Creation, transp., deformation of a 2500x2500 matrix (sec):
>> 0.8296667
>> 2400x2400 normal distributed random matrix ^1000 (sec):
>> 0.1553334
>> Sorting of 7,000,000 random values__ (sec):
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Hello,
Just out of interest, I ran benchmark-25.R from Simon’s repo, as I have access
to an M1 Max. Are the *very* long times on cross-product, linear regression,
and Matrix functions a consequence of the BLAS version?
Kieran
> sessionInfo()
R version 4.1.1 (2021-08-10)
Platform: aarch64-apple-darwin20 (64-bit)
Running under: macOS Monterey 12.0.1
Matrix products: default
BLAS:
/Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRblas.0.dylib
LAPACK:
/Library/Frameworks/R.framework/Versions/4.1-arm64/Resources/lib/libRlapack.dylib
> source("R-benchmark-25.R")
Loading required package: SuppDists
R Benchmark 2.5
===
Number of times each test is run__: 3
I. Matrix calculation
-
Creation, transp., deformation of a 2500x2500 matrix (sec): 0.2496656
2400x2400 normal distributed random matrix ^1000 (sec): 0.1050009
Sorting of 7,000,000 random values__ (sec): 0.5946673
2800x2800 cross-product matrix (b = a' * a)_ (sec): 13.30167
Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): 6.270334
Trimmed geom. mean (2 extremes eliminated): 0.976431082297569
II. Matrix functions
FFT over 2,400,000 random values (sec): 0.07266631
Eigenvalues of a 640x640 random matrix__ (sec): 0.4256672
Determinant of a 2500x2500 random matrix (sec): 1.738333
Cholesky decomposition of a 3000x3000 matrix (sec): 5.17
Inverse of a 1600x1600 random matrix (sec): 1.430996
Trimmed geom. mean (2 extremes eliminated): 1.01925013610031
III. Programmation
--
3,500,000 Fibonacci numbers calculation (vector calc)(sec): 0.09500273
Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): 0.1153335
Grand common divisors of 400,000 pairs (recursion)__ (sec): 0.07999841
Creation of a 500x500 Toeplitz matrix (loops)___ (sec): 0.01733593
Escoufier's method on a 45x45 matrix (mixed) (sec): 0.1529963
Trimmed geom. mean (2 extremes eliminated): 0.0957023962714685
Total time for all 15 tests_ (sec): 29.823
Overall mean (sum of I, II and III trimmed means/3)_ (sec): 0.45668322781674
--- End of test ---
Warning messages:
1: In remove("a", "b") : object 'a' not found
2: In remove("a", "b") : object 'b' not found
> On Oct 31, 2021, at 5:11 PM, Simon Urbanek
> wrote:
>
>
> Tim,
>
> that is a great idea, those test are really old. Just for the fun of it I
> have run the tests on my old iMac, but with R 4.1.2 and they still work.
> It's nice to see the huge speed improvements in loops and similar (see below
> - recall the original tests were scaled to be around 1).
>
> I have added the page to the repo
> https://github.com/R-macos/R-mac-dev
> so I'd be happy to review PRs, but I'll probably want to re-do it first so it
> is better organized for comparisons as we have to also accommodate M1 etc.
>
> Cheers,
> Simon
>
> ---
> iMac14,2 3.2Ghz i5, macOS 10.4.6, R 4.1.2 vecib/Accelerate BLAS
>
>
> R Benchmark 2.5
> ===
> Number of times each test is run__: 3
>
> I. Matrix calculation
> -
> Creation, transp., deformation of a 2500x2500 matrix (sec):
> 0.8296667
> 2400x2400 normal distributed random matrix ^1000 (sec):
> 0.1553334
> Sorting of 7,000,000 random values__ (sec):
> 0.6383334
> 2800x2800 cross-product matrix (b = a' * a)_ (sec):
> 0.2420001
> Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec):
> 0.170
>
> Trimmed geom. mean (2 extremes eliminated): 0.29781941072597
>
> II. Matrix functions
>
> FFT over 2,400,000 random values (sec):
> 0.331
> Eigenvalues of a 640x640 random matrix__ (sec):
> 0.3470001
> Determinant of a 2500x2500 random matrix (sec):
> 0.2070001
> Cholesky decomposition of a 3000x3000 matrix (sec):
> 0.2543334
> Inverse of a 1600x1600 random matrix (sec):
> 0.3456663
>
>Trimmed geom. mean (2 extremes eliminated): 0.307686639256803
>
> III. Programmation
> --
> 3,500,000 Fibonacci numbers calculation (vector calc)(sec): 0.245
> Creation of
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
I ran some tests using my packages and a bootstrapped linear model, and the results were that my M1 MacBook Air was faster than my iMac 8-core 2020, and that included using 4 parallel cores on the M1 versus 8 on the iMac. Ken > On 1 Nov 2021, at 8:11 am, Simon Urbanek wrote: > > > Tim, > > that is a great idea, those test are really old. Just for the fun of it I > have run the tests on my old iMac, but with R 4.1.2 and they still work. > It's nice to see the huge speed improvements in loops and similar (see below > - recall the original tests were scaled to be around 1). > > I have added the page to the repo > https://github.com/R-macos/R-mac-dev > so I'd be happy to review PRs, but I'll probably want to re-do it first so it > is better organized for comparisons as we have to also accommodate M1 etc. > > Cheers, > Simon > > --- > iMac14,2 3.2Ghz i5, macOS 10.4.6, R 4.1.2 vecib/Accelerate BLAS > > > R Benchmark 2.5 > === > Number of times each test is run__: 3 > > I. Matrix calculation > - > Creation, transp., deformation of a 2500x2500 matrix (sec): > 0.8296667 > 2400x2400 normal distributed random matrix ^1000 (sec): > 0.1553334 > Sorting of 7,000,000 random values__ (sec): > 0.6383334 > 2800x2800 cross-product matrix (b = a' * a)_ (sec): > 0.2420001 > Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): > 0.170 > > Trimmed geom. mean (2 extremes eliminated): 0.29781941072597 > > II. Matrix functions > > FFT over 2,400,000 random values (sec): > 0.331 > Eigenvalues of a 640x640 random matrix__ (sec): > 0.3470001 > Determinant of a 2500x2500 random matrix (sec): > 0.2070001 > Cholesky decomposition of a 3000x3000 matrix (sec): > 0.2543334 > Inverse of a 1600x1600 random matrix (sec): > 0.3456663 > >Trimmed geom. mean (2 extremes eliminated): 0.307686639256803 > > III. Programmation > -- > 3,500,000 Fibonacci numbers calculation (vector calc)(sec): 0.245 > Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): > 0.2896669 > Grand common divisors of 400,000 pairs (recursion)__ (sec): > 0.2593331 > Creation of a 500x500 Toeplitz matrix (loops)___ (sec): > 0.0415 > Escoufier's method on a 45x45 matrix (mixed) (sec): > 0.2630005 > >Trimmed geom. mean (2 extremes eliminated): 0.255658395143118 > > > Total time for all 15 tests_ (sec): 4.618667 > Overall mean (sum of I, II and III trimmed means/3)_ (sec): > 0.286136920519432 > --- End of test --- > > > >> On Nov 1, 2021, at 2:48 AM, Tim Bates wrote: >> >> I wonder if this (2008/R 2.7) page could be updated with some current >> benchmark runs? >> >> Especially current Intel server chips, i9, and M1/+ >> >> I'm guessing if Simon could help upload the resulting updated page, people >> here could contribute bench mark runs on different hardware. >> >> >> Also be interesting to see different blas results. >> >> I wonder if either intel or arm chip "neural" cores (dot product engines?) >> or multi-core and GPU are being used in current R builds? >> >> tim >> >> >> >> >> >> >> >> ___ >> R-SIG-Mac mailing list >> R-SIG-Mac@r-project.org >> https://stat.ethz.ch/mailman/listinfo/r-sig-mac >> > > ___ > R-SIG-Mac mailing list > R-SIG-Mac@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-mac ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
Re: [R-SIG-Mac] https://mac.r-project.org/benchmarks/
Tim, that is a great idea, those test are really old. Just for the fun of it I have run the tests on my old iMac, but with R 4.1.2 and they still work. It's nice to see the huge speed improvements in loops and similar (see below - recall the original tests were scaled to be around 1). I have added the page to the repo https://github.com/R-macos/R-mac-dev so I'd be happy to review PRs, but I'll probably want to re-do it first so it is better organized for comparisons as we have to also accommodate M1 etc. Cheers, Simon --- iMac14,2 3.2Ghz i5, macOS 10.4.6, R 4.1.2 vecib/Accelerate BLAS R Benchmark 2.5 === Number of times each test is run__: 3 I. Matrix calculation - Creation, transp., deformation of a 2500x2500 matrix (sec): 0.8296667 2400x2400 normal distributed random matrix ^1000 (sec): 0.1553334 Sorting of 7,000,000 random values__ (sec): 0.6383334 2800x2800 cross-product matrix (b = a' * a)_ (sec): 0.2420001 Linear regr. over a 3000x3000 matrix (c = a \ b')___ (sec): 0.170 Trimmed geom. mean (2 extremes eliminated): 0.29781941072597 II. Matrix functions FFT over 2,400,000 random values (sec): 0.331 Eigenvalues of a 640x640 random matrix__ (sec): 0.3470001 Determinant of a 2500x2500 random matrix (sec): 0.2070001 Cholesky decomposition of a 3000x3000 matrix (sec): 0.2543334 Inverse of a 1600x1600 random matrix (sec): 0.3456663 Trimmed geom. mean (2 extremes eliminated): 0.307686639256803 III. Programmation -- 3,500,000 Fibonacci numbers calculation (vector calc)(sec): 0.245 Creation of a 3000x3000 Hilbert matrix (matrix calc) (sec): 0.2896669 Grand common divisors of 400,000 pairs (recursion)__ (sec): 0.2593331 Creation of a 500x500 Toeplitz matrix (loops)___ (sec): 0.0415 Escoufier's method on a 45x45 matrix (mixed) (sec): 0.2630005 Trimmed geom. mean (2 extremes eliminated): 0.255658395143118 Total time for all 15 tests_ (sec): 4.618667 Overall mean (sum of I, II and III trimmed means/3)_ (sec): 0.286136920519432 --- End of test --- > On Nov 1, 2021, at 2:48 AM, Tim Bates wrote: > > I wonder if this (2008/R 2.7) page could be updated with some current > benchmark runs? > > Especially current Intel server chips, i9, and M1/+ > > I'm guessing if Simon could help upload the resulting updated page, people > here could contribute bench mark runs on different hardware. > > > Also be interesting to see different blas results. > > I wonder if either intel or arm chip "neural" cores (dot product engines?) or > multi-core and GPU are being used in current R builds? > > tim > > > > > > > > ___ > R-SIG-Mac mailing list > R-SIG-Mac@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-mac > ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
[R-SIG-Mac] https://mac.r-project.org/benchmarks/
I wonder if this (2008/R 2.7) page could be updated with some current benchmark runs? Especially current Intel server chips, i9, and M1/+ I'm guessing if Simon could help upload the resulting updated page, people here could contribute bench mark runs on different hardware. Also be interesting to see different blas results. I wonder if either intel or arm chip "neural" cores (dot product engines?) or multi-core and GPU are being used in current R builds? tim ___ R-SIG-Mac mailing list R-SIG-Mac@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mac
