Fused GPU bonded kernels

[alexxy/gromacs.git] / docs / release-notes / 2020 / major / performance.rst

diff --git a/docs/release-notes/2020/major/performance.rst b/docs/release-notes/2020/major/performance.rst
index b7ad1cd9317a5b628a0ba8784fd53e2bc02d64cd..9d8b32ef64584c19a8d34d395398c33ae7175770 100644 (file)
--- a/docs/release-notes/2020/major/performance.rst
+++ b/docs/release-notes/2020/major/performance.rst
@@ -16,3 +16,10 @@ if the (often noisy) FFTW auto-tuner picks an AVX512 kernel in a run that otherw
 only uses AVX/AVX2 which could run at higher CPU clocks without AVX512 clock speed limitation.
 Now AVX512 is only used for the internal FFTW if GROMACS is also configured with
 the same SIMD flavor.
+
+Bonded kernels on GPU have been fused
+"""""""""""""""""""""""""""""""""""""
+
+Instead of launching one GPU kernel for each listed interaction type there is now one
+GPU kernel that handles all listed interactions. This improves the performance when
+running bonded calculations on a GPU.