Skip to content

Benchmarks

Jump to bottom
Fabian Wein edited this page May 28, 2026 · 14 revisions

This is a fun page: collect measurement of your machine by using benchmark.py from share/python. Just run bechmark.py <your_cfs_binary>

  • MacBook Pro 2018 i5
# host:Fabians-Flunder mesh:30 comment:macbook_2018_i5_clang
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL_NUM_THREADS 7:res 8:comment 
cholmod 	34.231 	34.0462 	1 	1 	1 	30 	macbook_2018_i5_clang 	
cholmod 	19.62 	60.9211 	4 	4 	4 	30 	macbook_2018_i5_clang 	
shapemap 	89.604 	89.4626 	1 	1 	1 	30 	macbook_2018_i5_clang 	
shapemap 	27.282 	100.271 	4 	4 	4 	30 	macbook_2018_i5_clang
  • Desktop Intel(R) Xeon(R) CPU E3-1275 v5 @ 3.60GHz (2015)
# host:mojo mesh:30 comment:mojo using mkl
# host:mojo mesh:30 comment:default
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment 
cholmod 	26.937 	26.8892 	1 	1 	1 	30 	default 	
cholmod 	15.209 	48.4572 	4 	4 	4 	30 	default 	
pardiso 	35.996 	35.9632 	1 	1 	1 	30 	default 	
pardiso 	14.89 	54.8077 	4 	4 	4 	30 	default 
shapemap 	68.74 	68.7128 	1 	1 	1 	30 	mojo 	
shapemap 	20.275 	75.2065 	4 	4 	4 	30 	mojo

-Desktop Core(TM) i7-10700 CPU (2021) Windows - icx 2023.2

# host:mosim mesh:30 comment:default
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment
cholmod         25.542  25.543  1       1       1       30      default
cholmod         14.201  14.2    4       4       4       30      default
pardiso         33.519  33.52   1       1       1       30      default
pardiso         12.608  12.608  4       4       4       30      default
shapemap        62.422  62.422  1       1       1       30      default
shapemap        19.189  19.189  4       4       4       30      default
  • MacBook Pro 2021 M1 Pro (arm64) -> x86_64 binary via Rosetta 2
# host:Finsternis mesh:30 comment:m1_intel
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL_NUM_THREADS 7:res 8:comment 
cholmod 	47.035 	47.0285 	1 	1 	1 	30 	m1_intel 	
cholmod 	19.344 	67.3188 	4 	4 	4 	30 	m1_intel 	
shapemap 	69.692 	69.6888 	1 	1 	1 	30 	m1_intel 	
shapemap 	20.134 	75.169  	4 	4 	4 	30 	m1_inte
  • MacBook Pro 2021 M1 Pro (arm64) -> arm64 binary, cholmod from SuiteSparse 6.0 and openblas
# host:Finsternis mesh:30 comment: m1_arm
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL_NUM_THREADS 7:res 8:comment 
cholmod 	25.999 	25.9815 	1 	1 	1 	30 	m1_arm 	
cholmod 	12.987 	64.0969 	4 	4 	4 	30 	m1_arm 	
shapemap 	53.39 	53.3442 	1 	1 	1 	30 	m1_arm 	
shapemap 	16.52 	60.6482 	4 	4 	4 	30 	m1_arm
  • MacBook Pro 2021 M1 Pro (arm64) -> arm64 binary, Apple's accelerate and SuiteSparse 7
# host:Finsternis mesh:30 comment: m1_arm
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment 
cholmod 	13.623 	13.6198 	1 	1 	1 	30 	default 	
cholmod 	10.531 	16.2446 	4 	4 	4 	30 	default 	
shapemap 	53.303 	53.2751 	1 	1 	1 	30 	default 	
shapemap 	15.954 	56.1159 	4 	4 	4 	30 	default
  • MacBook Pro M5 Pro (arm64) -> arm64 binary, Apple's accelerate and SuiteSparse 7
# host:mocca mesh:30 comment:-
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:mem in GB 
shapemap 	 18.55 	 18.54 	1 	1 	1 	30 	0.3516 	
shapemap 	  6.10 	 19.88 	4 	4 	4 	30 	0.3545 	
cholmod  	  6.20 	  6.19 	1 	1 	1 	30 	1.2246 	
cholmod  	  5.82 	  6.68 	4 	4 	4 	30 	1.2266 	
ginkgo   	  8.64 	  8.63 	1 	1 	1 	30 	0.3652 	
ginkgo   	  3.14 	  9.86 	4 	4 	4 	30 	0.3613
  • Mac Studio 2023 M2 Ultra (arm64) -> arm64 binary, Apple's accelerate and SuiteSparse 7
# host:armin mesh:30 comment:default
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment 
cholmod 	10.858 	10.8612 	1 	1 	1 	30 	default 	
cholmod 	7.997 	12.5465 	4 	4 	4 	30 	default 	
shapemap 	48.634 	48.5678 	1 	1 	1 	30 	default 	
shapemap 	15.238 	52.1936 	4 	4 	4 	30 	default
  • AMD EPYC 7F32 8-Core Processor (2020)
# host:flash mesh:30 comment:flash
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL_NUM_THREADS 7:res 8:comment 
cholmod 	26.163 	26.1045 	1 	1 	1 	30 	flash 	
cholmod 	16.219 	55.8605 	4 	4 	4 	30 	flash 	
shapemap 	59.942 	59.9081 	1 	1 	1 	30 	flash 	
shapemap 	18.442 	67.3598 	4 	4 	4 	30 	flash
  • AMD EPYC 7F32 8-Core Processor (2020) oneAPI 2025
# host:flash mesh:30 comment:default
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment 
cholmod 	33.122 	33.0725 	1 	1 	1 	30 	default 	
cholmod 	24.75 	75.7349 	4 	4 	4 	30 	default 	
pardiso 	39.721 	39.6672 	1 	1 	1 	30 	default 	
pardiso 	15.905 	52.3382 	4 	4 	4 	30 	default 	
shapemap 	61.265 	61.2309 	1 	1 	1 	30 	default 	
shapemap 	18.844 	62.6649 	4 	4 	4 	30 	defaul
  • AMD EPYC 7402 24-Core Processor (2020)
# host:neubau-11 mesh:30 comment:manni
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL_NUM_THREADS 7:res 8:comment 
cholmod 	30.796 	30.7417 	1 	1 	1 	30 	manni 	
cholmod 	18.201 	59.3964 	4 	4 	4 	30 	manni 	
shapemap 	70.13 	70.0903 	1 	1 	1 	30 	manni 	
shapemap 	19.718 	74.9371 	4 	4 	4 	30 	manni
  • Intel(R) Core(TM) i9-14900KF 24-Core Processor (2025)
# host:shady mesh:30 comment:default
# 1:test 2:wall 3:cpu 4:CFS_NUM_THREADS 5:OMP_NUM_THREADS 6:MKL/VECLIB_THREADS 7:res 8:comment 
cholmod 	15.404 	15.3537 	1 	1 	1 	30 	default 	
cholmod 	9.63 	21.0741 	4 	4 	4 	30 	default 	
pardiso 	19.593 	19.5532 	1 	1 	1 	30 	default 	
pardiso 	6.779 	22.043 	4 	4 	4 	30 	default 	
shapemap 	21.119 	21.1121 	1 	1 	1 	30 	default 	
shapemap 	7.145 	22.2056 	4 	4 	4 	30 	default

Clone this wiki locally