Skip to content

Use nano-gemm instead of matrixmultiply #292

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
cschwan wants to merge 2 commits into from
Closed

Use nano-gemm instead of matrixmultiply #292

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a4e96a2
Use `nano-gemm` instead of `matrixmultiply` for non-DIS grids
cschwan Jun 6, 2024
d20fbb9
Merge branch 'master' into test-nano-gemm
cschwan Jun 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
92 changes: 88 additions & 4 deletions Cargo.lock
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

1 change: 1 addition & 0 deletions pineappl/Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,7 @@ float-cmp = "0.9.0"
git-version = "0.3.5"
itertools = "0.10.1"
lz4_flex = "0.9.2"
nano-gemm = { git = "https://github.com/sarah-ek/nano-gemm", package = "nano-gemm" }
ndarray = { features = ["serde"], version = "0.15.4" }
rustc-hash = "1.1.0"
serde = { features = ["derive"], version = "1.0.130" }
Expand Down
40 changes: 36 additions & 4 deletions pineappl/src/evolution.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,8 +10,7 @@ use super::sparse_array3::SparseArray3;
use super::subgrid::{Mu2, Subgrid, SubgridEnum};
use float_cmp::approx_eq;
use itertools::Itertools;
use ndarray::linalg;
use ndarray::{s, Array1, Array2, Array3, ArrayView1, ArrayView4, Axis};
use ndarray::{s, Array1, Array2, Array3, ArrayView1, ArrayView2, ArrayView4, Axis};
use std::iter;

/// Number of ULPS used to de-duplicate grid values in [`Grid::evolve_info`].
Expand Down Expand Up @@ -170,6 +169,39 @@ fn gluon_has_pid_zero(grid: &Grid) -> bool {
&& grid.pid_basis() == PidBasis::Pdg
}

fn nano_gemm_mat_mul(
alpha: f64,
a: ArrayView2<f64>,
b: ArrayView2<f64>,
beta: f64,
c: &mut Array2<f64>,
) {
use nano_gemm::planless;

let ((m, k), (_, n)) = (a.dim(), b.dim());

unsafe {
planless::execute_f64(
m,
n,
k,
c.as_mut_ptr(),
c.strides()[0],
c.strides()[1],
a.as_ptr(),
a.strides()[0],
a.strides()[1],
b.as_ptr(),
b.strides()[0],
b.strides()[1],
beta,
alpha,
false,
false,
);
}
}

type Pid01IndexTuples = Vec<(usize, usize)>;
type Pid01Tuples = Vec<(i32, i32)>;

Expand Down Expand Up @@ -602,8 +634,8 @@ pub(crate) fn evolve_slice_with_two(
.map(|(opa, opb)| (fk_table, opa, opb))
},
) {
linalg::general_mat_mul(1.0, &array, &opb.t(), 0.0, &mut tmp);
linalg::general_mat_mul(factor, opa, &tmp, 1.0, fk_table);
nano_gemm_mat_mul(1.0, array.view(), opb.t(), 0.0, &mut tmp);
nano_gemm_mat_mul(factor, opa.view(), tmp.view(), 1.0, fk_table);
}
}
}
Expand Down