Benchmarks

Here are some simple benchmarks. Take them with a grain of salt since they run on virtual machines in the cloud to generate the documentation automatically.

First-derivative operators

Periodic domains

Let's set up some benchmark code.

using BenchmarkToolsusing LinearAlgebra, SparseArraysusing SummationByPartsOperatorsBLAS.set_num_threads(1) # make sure that BLAS is serial to be fairT = Float64xmin, xmax = T(0), T(1)D_SBP = periodic_derivative_operator(derivative_order=1, accuracy_order=2,                                     xmin=xmin, xmax=xmax, N=100)x = grid(D_SBP)D_sparse = sparse(D_SBP)u = randn(eltype(D_SBP), length(x)); du = similar(u);@show D_SBP * u  D_sparse * ufunction doit(D, text, du, u)  println(text)  sleep(0.1)  show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D, $u))  println()end
doit (generic function with 1 method)

First, we benchmark the implementation from SummationByPartsOperators.jl.

doit(D_SBP, "D_SBP:", du, u)
D_SBP:
BenchmarkTools.Trial: 10000 samples with 996 evaluations per sample.
 Range (minmax):  25.791 ns49.279 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     26.545 ns               GC (median):    0.00%
 Time  (mean ± σ):   26.850 ns ±  1.592 ns   GC (mean ± σ):  0.00% ± 0.00%

   ▄█▁▂                                                     
  ▃████▆▃▄▃▃▃▃▂▂▂▂▂▂▂▂▁▂▁▂▁▂▁▂▂▁▂▁▁▂▂▂▁▂▂▂▁▁▁▁▂▂▁▁▂▂▂▂▂▂▂▂ ▃
  25.8 ns         Histogram: frequency by time        34.9 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.

Next, we compare this to the runtime obtained using a sparse matrix representation of the derivative operator. Depending on the hardware etc., this can be an order of magnitude slower than the optimized implementation from SummationByPartsOperators.jl.

doit(D_sparse, "D_sparse:", du, u)
D_sparse:
BenchmarkTools.Trial: 10000 samples with 585 evaluations per sample.
 Range (minmax):  201.369 ns475.164 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     212.415 ns                GC (median):    0.00%
 Time  (mean ± σ):   214.598 ns ±  10.600 ns   GC (mean ± σ):  0.00% ± 0.00%

            ▁▂▅▆▆██▇▅▃▁                                       
  ▁▁▁▁▁▂▃▄▅▆███████████▇▆▅▄▄▃▃▃▃▂▃▃▃▄▃▃▄▃▃▃▃▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁ ▃
  201 ns           Histogram: frequency by time          239 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.

These results were obtained using the following versions.

using InteractiveUtilsversioninfo()using PkgPkg.status(["SummationByPartsOperators"],           mode=PKGMODE_MANIFEST)
Julia Version 1.10.12
Commit d93beab124c (2026-08-15 10:29 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver3)
Threads: 2 default, 0 interactive, 1 GC (on 4 virtual cores)
Environment:
  JULIA_PKG_SERVER_REGISTRY_PREFERENCE = eager
Status `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl/docs/Manifest.toml`
  [9f78cca6] SummationByPartsOperators v0.5.101 `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl`

Bounded domains

We start again by setting up some benchmark code.

using BenchmarkToolsusing LinearAlgebra, SparseArraysusing SummationByPartsOperators, BandedMatricesBLAS.set_num_threads(1) # make sure that BLAS is serial to be fairT = Float64xmin, xmax = T(0), T(1)D_SBP = derivative_operator(MattssonNordström2004(), derivative_order=1,                            accuracy_order=6, xmin=xmin, xmax=xmax, N=10^3)D_sparse = sparse(D_SBP)D_banded = BandedMatrix(D_SBP)u = randn(eltype(D_SBP), size(D_SBP, 1)); du = similar(u);@show D_SBP * u  D_sparse * u  D_banded * ufunction doit(D, text, du, u)  println(text)  sleep(0.1)  show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D, $u))  println()end
doit (generic function with 1 method)

First, we benchmark the implementation from SummationByPartsOperators.jl.

doit(D_SBP, "D_SBP:", du, u)
D_SBP:
BenchmarkTools.Trial: 10000 samples with 203 evaluations per sample.
 Range (minmax):  383.133 ns761.177 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     388.611 ns                GC (median):    0.00%
 Time  (mean ± σ):   392.313 ns ±  13.194 ns   GC (mean ± σ):  0.00% ± 0.00%

      ▄█                                                       
  ▂▃▃▅██▇▄▃▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▁▁▂▁▂▁▁▁▁▂▂▁▁▁▂▂▂▂▂▂▃▃▃▂▂▂▂▂▂▂▂▂▂▂ ▃
  383 ns           Histogram: frequency by time          439 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.

Again, we compare this to a representation of the derivative operator as a sparse matrix. No surprise - it is again much slower, as in periodic domains.

doit(D_sparse, "D_sparse:", du, u)
D_sparse:
BenchmarkTools.Trial: 10000 samples with 7 evaluations per sample.
 Range (minmax):  4.474 μs 17.630 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     4.609 μs                GC (median):    0.00%
 Time  (mean ± σ):   4.701 μs ± 581.499 ns   GC (mean ± σ):  0.00% ± 0.00%

   ▃▆█▆▃                             ▁▂▂                    ▂
  ▆██████▄▄▆▆▅▅▅▄▁▁▁▁▅▃▄▅▁▄▅▄▄▅▄▄▃▃▃▆████▇▆▅▁▄▄▄▅▆▅▄▅▄▄▁▁▃▃ █
  4.47 μs      Histogram: log(frequency) by time      6.47 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Finally, we compare it to a representation as a banded matrix. Disappointingly, this is still much slower than the optimized implementation from SummationByPartsOperators.jl.

doit(D_banded, "D_banded:", du, u)
D_banded:
BenchmarkTools.Trial: 10000 samples with 1 evaluation per sample.
 Range (minmax):  9.497 μs32.040 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     9.518 μs               GC (median):    0.00%
 Time  (mean ± σ):   9.643 μs ±  1.016 μs   GC (mean ± σ):  0.00% ± 0.00%

   ▄▅▄█▆▁▁▁▁▃▁▃▃▄▁▄▃▃▁▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▃▁▁▃▁▁▁▁▁▁▁▃▁▁▁▁▃▄▆ █
  9.5 μs       Histogram: log(frequency) by time     17.6 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

These results were obtained using the following versions.

using InteractiveUtilsversioninfo()using PkgPkg.status(["SummationByPartsOperators", "BandedMatrices"],           mode=PKGMODE_MANIFEST)
Julia Version 1.10.12
Commit d93beab124c (2026-08-15 10:29 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver3)
Threads: 2 default, 0 interactive, 1 GC (on 4 virtual cores)
Environment:
  JULIA_PKG_SERVER_REGISTRY_PREFERENCE = eager
Status `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl/docs/Manifest.toml`
  [aae01518] BandedMatrices v1.12.0
  [9f78cca6] SummationByPartsOperators v0.5.101 `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl`

Dissipation operators

We follow the same structure as before. At first, we set up some benchmark code.

using BenchmarkToolsusing LinearAlgebra, SparseArraysusing SummationByPartsOperators, BandedMatricesBLAS.set_num_threads(1) # make sure that BLAS is serial to be fairT = Float64xmin, xmax = T(0), T(1)D_SBP = derivative_operator(MattssonNordström2004(), derivative_order=1,                            accuracy_order=6, xmin=xmin, xmax=xmax, N=10^3)Di_SBP  = dissipation_operator(MattssonSvärdNordström2004(), D_SBP)Di_sparse = sparse(Di_SBP)Di_banded = BandedMatrix(Di_SBP)Di_full   = Matrix(Di_SBP)u = randn(eltype(D_SBP), size(D_SBP, 1)); du = similar(u);@show Di_SBP * u  Di_sparse * u  Di_banded * u  Di_full * ufunction doit(D, text, du, u)  println(text)  sleep(0.1)  show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D, $u))  println()end
doit (generic function with 1 method)

At first, let us benchmark the derivative and dissipation operators implemented in SummationByPartsOperators.jl.

doit(D_SBP, "D_SBP:", du, u)doit(Di_SBP, "Di_SBP:", du, u)
D_SBP:
BenchmarkTools.Trial: 10000 samples with 203 evaluations per sample.
 Range (minmax):  383.182 ns530.355 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     386.143 ns                GC (median):    0.00%
 Time  (mean ± σ):   389.835 ns ±  12.281 ns   GC (mean ± σ):  0.00% ± 0.00%

  ▃▇██▆▄▄▂                                       ▂▃▃▂▁        ▂
  █████████▆▅▆▇▇▇▆▅▃▄▁▃▁▁▁▁▁▃▃▄▁▁▁▁▃▁▃▃▁▁▁▃▃▃▁▃▁▇███████▇▇▆▅▅ █
  383 ns        Histogram: log(frequency) by time        434 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
Di_SBP:
BenchmarkTools.Trial: 10000 samples with 16 evaluations per sample.
 Range (minmax):  984.938 ns 2.253 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     993.125 ns               GC (median):    0.00%
 Time  (mean ± σ):     1.004 μs ± 71.342 ns   GC (mean ± σ):  0.00% ± 0.00%

  █                                                         ▁ ▂
  █▄▁▃█▆▁▃▁▁▁▃▁▆▄▄▄▁▄▄▃▄▅▄▄▃▄▅▁▄▁▃▁▁▃▁▄▄▁▁▁▁▁▃▁▄▁▁▁▁▁▃▁▁▃▁▅█ █
  985 ns        Histogram: log(frequency) by time       1.5 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Next, we compare the results to sparse matrix representations. It will not come as a surprise that these are again much (around an order of magnitude) slower.

doit(Di_sparse, "Di_sparse:", du, u)doit(Di_banded, "Di_banded:", du, u)
Di_sparse:
BenchmarkTools.Trial: 10000 samples with 6 evaluations per sample.
 Range (minmax):  5.016 μs 10.159 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     5.091 μs                GC (median):    0.00%
 Time  (mean ± σ):   5.141 μs ± 273.065 ns   GC (mean ± σ):  0.00% ± 0.00%

  ▂▆█ ▁▁                                               ▁▁▁ ▂
  ███▇███▅▁▁▁▁▅▆▅▆▃▄▁▁▁▁▁▁▁▁▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▃▁▁▁▁▁▁▁▁▁▇███ █
  5.02 μs      Histogram: log(frequency) by time       6.5 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.
Di_banded:
BenchmarkTools.Trial: 10000 samples with 5 evaluations per sample.
 Range (minmax):  6.891 μs 13.974 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     6.949 μs                GC (median):    0.00%
 Time  (mean ± σ):   7.024 μs ± 401.688 ns   GC (mean ± σ):  0.00% ± 0.00%

  ▃█  ▁▁                                               ▁▁   ▂
  ██▅██▄▁▁▁▁▁▃▅▆▅▄▅▃▁▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▄████▇ █
  6.89 μs      Histogram: log(frequency) by time      8.69 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Finally, let's benchmark the same computation if a full (dense) matrix is used to represent the derivative operator. This is obviously a bad idea but 🤷

doit(Di_full, "Di_full:", du, u)
Di_full:
BenchmarkTools.Trial: 10000 samples with 1 evaluation per sample.
 Range (minmax):  117.730 μs199.134 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     148.288 μs                GC (median):    0.00%
 Time  (mean ± σ):   150.016 μs ±   5.911 μs   GC (mean ± σ):  0.00% ± 0.00%

                            ▁▁▁▂▂▄▄▆▇█▄▃▂▂▁▁▂▃▄▄▅▅▃▂▁        ▃
  ▇▆▁▁▄▃▁▁▁▄▁▃▁▃▄▄▃▃▁▁▁▁▁▁▃▇████████████████████████████▇█▇█ █
  118 μs        Histogram: log(frequency) by time        169 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

These results were obtained using the following versions.

using InteractiveUtilsversioninfo()using PkgPkg.status(["SummationByPartsOperators", "BandedMatrices"],           mode=PKGMODE_MANIFEST)
Julia Version 1.10.12
Commit d93beab124c (2026-08-15 10:29 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver3)
Threads: 2 default, 0 interactive, 1 GC (on 4 virtual cores)
Environment:
  JULIA_PKG_SERVER_REGISTRY_PREFERENCE = eager
Status `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl/docs/Manifest.toml`
  [aae01518] BandedMatrices v1.12.0
  [9f78cca6] SummationByPartsOperators v0.5.101 `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl`

Structure-of-Arrays (SoA) and Array-of-Structures (AoS)

SummationByPartsOperators.jl tries to provide efficient support of

To demonstrate this, let us set up some benchmark code.

using BenchmarkToolsusing StaticArrays, StructArraysusing LinearAlgebra, SparseArraysusing SummationByPartsOperatorsBLAS.set_num_threads(1) # make sure that BLAS is serial to be fairstruct Vec5{T} <: FieldVector{5,T}  x1::T  x2::T  x3::T  x4::T  x5::Tend# Apply `mul!` to each component of a plain array of structures one after anotherfunction mul_aos!(du, D, u, args...)  for i in 1:size(du, 1)    mul!(view(du, i, :), D, view(u, i, :), args...)  endendT = Float64xmin, xmax = T(0), T(1)D_SBP = derivative_operator(MattssonNordström2004(), derivative_order=1,                            accuracy_order=4, xmin=xmin, xmax=xmax, N=101)D_sparse = sparse(D_SBP)D_full   = Matrix(D_SBP)
101×101 Matrix{Float64}:
 -141.176    173.529   -23.5294   …    0.0         0.0       0.0
  -50.0        0.0      50.0           0.0         0.0       0.0
    9.30233  -68.6047    0.0           0.0         0.0       0.0
    3.06122    0.0     -60.2041        0.0         0.0       0.0
    0.0        0.0       8.33333       0.0         0.0       0.0
    0.0        0.0       0.0      …    0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    ⋮                             ⋱                          ⋮
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0           0.0         0.0       0.0
    0.0        0.0       0.0      …    0.0         0.0       0.0
    0.0        0.0       0.0          -8.33333     0.0       0.0
    0.0        0.0       0.0          60.2041      0.0      -3.06122
    0.0        0.0       0.0           0.0        68.6047   -9.30233
    0.0        0.0       0.0         -50.0         0.0      50.0
    0.0        0.0       0.0      …   23.5294   -173.529   141.176

At first, we benchmark the application of the operators implemented in SummationByPartsOperators.jl and their representations as sparse and dense matrices in the scalar case. As before, the sparse matrix representation is around an order of magnitude slower and the dense matrix representation is far off.

println("Scalar case")u = randn(T, size(D_SBP, 1)); du = similar(u)println("D_SBP")show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D_SBP, $u))println("\nD_sparse")show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D_sparse, $u))println("\nD_full")show(stdout, MIME"text/plain"(), @benchmark mul!($du, $D_full, $u))
Scalar case
D_SBP
BenchmarkTools.Trial: 10000 samples with 988 evaluations per sample.
 Range (minmax):  46.656 ns79.623 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     48.056 ns               GC (median):    0.00%
 Time  (mean ± σ):   48.579 ns ±  2.175 ns   GC (mean ± σ):  0.00% ± 0.00%

      ▃█▇▅▃▁                                                 
  ▂▂▃▅██████▅▄▃▃▂▂▃▂▂▂▂▂▂▂▁▂▁▁▁▁▁▁▂▁▂▂▂▂▁▁▁▁▁▂▂▂▂▂▃▂▂▂▂▂▂▂▂ ▃
  46.7 ns         Histogram: frequency by time        57.4 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_sparse
BenchmarkTools.Trial: 10000 samples with 224 evaluations per sample.
 Range (minmax):  326.951 ns595.804 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     390.326 ns                GC (median):    0.00%
 Time  (mean ± σ):   391.388 ns ±  22.592 ns   GC (mean ± σ):  0.00% ± 0.00%

                        ▂▃▃▂▅▆▆█▆▅▃▁▁                         
  ▁▂▂▂▂▁▁▁▁▂▃▃▄▅▄▄▄▄▄▅▅██████████████▇▆▆▅▅▅▅▄▄▃▃▃▃▃▃▂▂▂▂▂▂▁▂▁ ▄
  327 ns           Histogram: frequency by time          455 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_full
BenchmarkTools.Trial: 10000 samples with 10 evaluations per sample.
 Range (minmax):  1.100 μs 2.751 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     1.112 μs               GC (median):    0.00%
 Time  (mean ± σ):   1.124 μs ± 99.502 ns   GC (mean ± σ):  0.00% ± 0.00%

  █                                                        ▁
  █▅▄▆▇▃▃▁▃▁▁▁▁▁▄▁▃▁▁▁▁▁▁▁▁▁▁▃▃▁▃▁▁▁▃▁▁▁▃▁▃▃▁▄▃▃▁▄▁▁▁▁▁▁▃▆ █
  1.1 μs       Histogram: log(frequency) by time     1.95 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Next, we use a plain array of structures (AoS) in the form of a two-dimensional array and our custom mul_aos! implementation that loops over each component, using mul! on views. Here, the differences between the timings are less pronounced.

println("Plain Array of Structures")u_aos_plain = randn(T, 5, size(D_SBP, 1)); du_aos_plain = similar(u_aos_plain)println("D_SBP")show(stdout, MIME"text/plain"(), @benchmark mul_aos!($du_aos_plain, $D_SBP, $u_aos_plain))println("\nD_sparse")show(stdout, MIME"text/plain"(), @benchmark mul_aos!($du_aos_plain, $D_sparse, $u_aos_plain))println("\nD_full")show(stdout, MIME"text/plain"(), @benchmark mul_aos!($du_aos_plain, $D_full, $u_aos_plain))
Plain Array of Structures
D_SBP
BenchmarkTools.Trial: 10000 samples with 10 evaluations per sample.
 Range (minmax):  1.278 μs  3.937 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     1.285 μs                GC (median):    0.00%
 Time  (mean ± σ):   1.302 μs ± 107.286 ns   GC (mean ± σ):  0.00% ± 0.00%

  █    ▁                                                     ▁
  █▄▃▄█▄▁▁▁▁▁▁▁▅▅▃▁▃▃▃▁▄▃▄▃▃▁▄▁▄▄▄▆▅▄▆▄▄▄▃▁▄▁▁▃▁▁▁▁▃▃▁▁▁▁▄▇ █
  1.28 μs      Histogram: log(frequency) by time      2.08 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_sparse
BenchmarkTools.Trial: 10000 samples with 9 evaluations per sample.
 Range (minmax):  2.471 μs  6.490 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     2.574 μs                GC (median):    0.00%
 Time  (mean ± σ):   2.604 μs ± 174.986 ns   GC (mean ± σ):  0.00% ± 0.00%

     ▁▇█                                                    
  ▂▂▄███▇▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▁▂▁▂▁▂▂▁▂▁▂▁▁▁▁▁▁▁▁▁▂▁▂▂▂▂▂▂▂ ▃
  2.47 μs         Histogram: frequency by time        3.52 μs <

 Memory estimate: 240 bytes, allocs estimate: 5.
D_full
BenchmarkTools.Trial: 10000 samples with 5 evaluations per sample.
 Range (minmax):  6.250 μs 12.233 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     6.290 μs                GC (median):    0.00%
 Time  (mean ± σ):   6.358 μs ± 346.597 ns   GC (mean ± σ):  0.00% ± 0.00%

  ▇▅▂  ▁                                                 ▁▂  ▁
  ███▇██▅▄▃▃▁▁▄▅▆▅▄▄▄▁▃▁▁▁▁▃▁▁▁▁▁▁▁▁▃▁▁▁▁▁▃▁▁▁▁▁▁▁▃▁▁▁▁▆███ █
  6.25 μs      Histogram: log(frequency) by time      8.01 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Now, we use an array of structures (AoS) based on reinterpret and standard mul!. This is much more efficient for the implementation in SummationByPartsOperators.jl. In Julia v1.6, this is also more efficient for sparse matrices but less efficient for dense matrices (compared to the plain AoS approach with mul_aos! above).

println("Array of Structures (reinterpreted array)")u_aos_r = reinterpret(reshape, Vec5{T}, u_aos_plain); du_aos_r = similar(u_aos_r)@show D_SBP * u_aos_r  D_sparse * u_aos_r  D_full * u_aos_rmul!(du_aos_r, D_SBP, u_aos_r)@show reinterpret(reshape, T, du_aos_r)  du_aos_plainprintln("D_SBP")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos_r, $D_SBP, $u_aos_r))println("\nD_sparse")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos_r, $D_sparse, $u_aos_r))println("\nD_full")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos_r, $D_full, $u_aos_r))
Array of Structures (reinterpreted array)
D_SBP * u_aos_r ≈ D_sparse * u_aos_r ≈ D_full * u_aos_r = true
reinterpret(reshape, T, du_aos_r) ≈ du_aos_plain = true
D_SBP
BenchmarkTools.Trial: 10000 samples with 501 evaluations per sample.
 Range (minmax):  210.295 ns297.443 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     213.735 ns                GC (median):    0.00%
 Time  (mean ± σ):   215.676 ns ±   5.798 ns   GC (mean ± σ):  0.00% ± 0.00%

        ▅██                                                
  ▂▂▂▃▄█████▇▆▄▃▂▂▂▂▂▂▂▂▂▁▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▂▁▂▂▂▂▃▄▄▃▃▃▂▂▂▂▂▂▂▂ ▃
  210 ns           Histogram: frequency by time          234 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_sparse
BenchmarkTools.Trial: 10000 samples with 187 evaluations per sample.
 Range (minmax):  549.856 ns 11.819 μs   GC (min … max): 0.00% … 93.17%
 Time  (median):     568.556 ns                GC (median):    0.00%
 Time  (mean ± σ):   576.331 ns ± 115.243 ns   GC (mean ± σ):  0.19% ±  0.93%

          ▃▆██▄▂                                               
  ▁▁▁▂▂▄▅████████▆▅▃▃▃▃▂▂▂▁▁▁▁▁▁▁▁▁▁▂▂▂▃▃▃▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁ ▃
  550 ns           Histogram: frequency by time          642 ns <

 Memory estimate: 32 bytes, allocs estimate: 1.
D_full
BenchmarkTools.Trial: 10000 samples with 1 evaluation per sample.
 Range (minmax):  13.936 μs60.043 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     14.016 μs               GC (median):    0.00%
 Time  (mean ± σ):   14.239 μs ±  1.374 μs   GC (mean ± σ):  0.00% ± 0.00%

  █  ▂▂▂▂▂▁▂▂▂▂▁▂▂▂▂▂▂▂▂▂▂▂▁▁▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▂▁▂▂▂▂▂▂▁▂▂ ▂
  13.9 μs         Histogram: frequency by time        22.2 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Next, we still use an array of structures (AoS), but copy the data into a plain Array instead of using the reinterpreted versions. There is no significant difference to the previous version in this case.

println("Array of Structures")u_aos = Array(u_aos_r); du_aos = similar(u_aos)@show D_SBP * u_aos  D_sparse * u_aos  D_full * u_aosmul!(du_aos, D_SBP, u_aos)@show du_aos  du_aos_rprintln("D_SBP")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos, $D_SBP, $u_aos))println("\nD_sparse")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos, $D_sparse, $u_aos))println("\nD_full")show(stdout, MIME"text/plain"(), @benchmark mul!($du_aos, $D_full, $u_aos))
Array of Structures
D_SBP * u_aos ≈ D_sparse * u_aos ≈ D_full * u_aos = true
du_aos ≈ du_aos_r = true
D_SBP
BenchmarkTools.Trial: 10000 samples with 506 evaluations per sample.
 Range (minmax):  212.711 ns398.393 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     215.879 ns                GC (median):    0.00%
 Time  (mean ± σ):   218.055 ns ±   6.711 ns   GC (mean ± σ):  0.00% ± 0.00%

     ▁▄▇██▆▅▄▃                              ▂▃▄▄▃▂▁           ▃
  ▄▅███████████▇▇███▇▇▅▄▁▁▃▆▆▁▄▁▃▁▁▃▄▃▃▃▅▃▆▇█████████▇▇▇▇▇▅▆▅ █
  213 ns        Histogram: log(frequency) by time        237 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_sparse
BenchmarkTools.Trial: 10000 samples with 141 evaluations per sample.
 Range (minmax):  674.454 ns 1.043 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     732.504 ns               GC (median):    0.00%
 Time  (mean ± σ):   738.607 ns ± 23.175 ns   GC (mean ± σ):  0.00% ± 0.00%

                       ▁▃▇█▇                                
  ▂▁▁▂▁▂▁▁▂▁▂▂▁▂▂▂▂▃▄▅▇█████▇▅▅▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▃▃▄▄▄▃▃▃▃▃▂▂▂ ▃
  674 ns          Histogram: frequency by time          810 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_full
BenchmarkTools.Trial: 10000 samples with 1 evaluation per sample.
 Range (minmax):  13.895 μs42.580 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     13.966 μs               GC (median):    0.00%
 Time  (mean ± σ):   14.350 μs ±  1.793 μs   GC (mean ± σ):  0.00% ± 0.00%

   █▄▄█▆▁▄▄▃▁▁▁▄▅▄▄▄▄▄▄▄▄▄▅▃▄▄▄▆▅▃▅▅▅▅▅▄▃▄▇█▇▆▆▆▄▆▇▆▃▄▅▇▇▇▆▆ █
  13.9 μs      Histogram: log(frequency) by time      22.7 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

Finally, let's look at a structure of arrays (SoA). Interestingly, this is slower than the array of structures we used above. On Julia v1.6, the sparse matrix representation performs particularly bad in this case.

println("Structure of Arrays")u_soa = StructArray(u_aos); du_soa = similar(u_soa)@show D_SBP * u_soa  D_sparse * u_soa  D_full * u_soamul!(du_soa, D_SBP, u_soa)@show du_soa  du_aosprintln("D_SBP")show(stdout, MIME"text/plain"(), @benchmark mul!($du_soa, $D_SBP, $u_soa))println("\nD_sparse")show(stdout, MIME"text/plain"(), @benchmark mul!($du_soa, $D_sparse, $u_soa))println("\nD_full")show(stdout, MIME"text/plain"(), @benchmark mul!($du_soa, $D_full, $u_soa))
Structure of Arrays
D_SBP * u_soa ≈ D_sparse * u_soa ≈ D_full * u_soa = true
du_soa ≈ du_aos = true
D_SBP
BenchmarkTools.Trial: 10000 samples with 487 evaluations per sample.
 Range (minmax):  222.945 ns312.002 ns   GC (min … max): 0.00% … 0.00%
 Time  (median):     226.236 ns                GC (median):    0.00%
 Time  (mean ± σ):   228.339 ns ±   6.569 ns   GC (mean ± σ):  0.00% ± 0.00%

      ▂▆█▆                                                 
  ▂▃▅▇██████▇▅▄▃▃▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▂▂▂▂▃▃▄▄▃▃▃▃▂▃▂▂▂▂▂▂▂ ▃
  223 ns           Histogram: frequency by time          248 ns <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_sparse
BenchmarkTools.Trial: 10000 samples with 1 evaluation per sample.
 Range (minmax):  48.901 μs108.503 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     49.092 μs                GC (median):    0.00%
 Time  (mean ± σ):   49.644 μs ±   2.641 μs   GC (mean ± σ):  0.00% ± 0.00%

  ▆▅▁  ▂                                               ▂▂     ▁
  ███▅█▇▄▁▁▁▃▅▆▆▆▄▅▃▁▄▁▃▃▁▁▁▁▁▁▁▁▃▁▁▁▁▁▄▁▁▁▁▁▁▁▁▁▁▁▁▁▇███▇▆▆ █
  48.9 μs       Histogram: log(frequency) by time      57.8 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.
D_full
BenchmarkTools.Trial: 10000 samples with 6 evaluations per sample.
 Range (minmax):  5.570 μs 11.797 μs   GC (min … max): 0.00% … 0.00%
 Time  (median):     5.612 μs                GC (median):    0.00%
 Time  (mean ± σ):   5.677 μs ± 331.523 ns   GC (mean ± σ):  0.00% ± 0.00%

  ▅█   ▁                                            ▁▁       ▁
  ████▅▃▁▁▁▄▅▅▄▅▁▄▁▁▁▁▁▁▃▃▁▁▄▁▄▁▁▃▃▁▃▃▃▁▃▁▁▃▃▃▁▁▁▇███▇▅▅▄▄ █
  5.57 μs      Histogram: log(frequency) by time      7.17 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

These results were obtained using the following versions.

using InteractiveUtilsversioninfo()using PkgPkg.status(["SummationByPartsOperators", "StaticArrays", "StructArrays"],           mode=PKGMODE_MANIFEST)
Julia Version 1.10.12
Commit d93beab124c (2026-08-15 10:29 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, znver3)
Threads: 2 default, 0 interactive, 1 GC (on 4 virtual cores)
Environment:
  JULIA_PKG_SERVER_REGISTRY_PREFERENCE = eager
Status `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl/docs/Manifest.toml`
  [90137ffa] StaticArrays v1.9.22
  [09ab397b] StructArrays v0.7.3
  [9f78cca6] SummationByPartsOperators v0.5.101 `~/work/SummationByPartsOperators.jl/SummationByPartsOperators.jl`