arma-thesis

git clone https://git.igankevich.com/arma-thesis.git
Log | Files | Refs | LICENSE

commit 9268b9ecee63bdf8505827b5a62df0d548e1ddcf
parent 678ca41cfba97412ec6a603062fc555fee0836a5
Author: Ivan Gankevich <igankevich@ya.ru>
Date:   Wed,  9 Aug 2017 18:47:46 +0300

Fix export of org-mode code blocks.

Diffstat:
arma-thesis.org | 6++----
1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/arma-thesis.org b/arma-thesis.org @@ -3501,7 +3501,7 @@ dependencies between parts computed in parallel, but requires more computational power (floating point operations per seconds). **** Performance of OpenMP and OpenCL implementations. :PROPERTIES: -:header-args:R: :results output org +:header-args:R: :results output raw :exports results :END: Differences in models' parallel algorithms make them efficient on different @@ -3604,7 +3604,7 @@ complex information dependencies, non-intensive calculations (multiplications and additions) and complex memory access patterns. #+name: tab-arma-performance -#+begin_src R :results output org :exports results +#+begin_src R source(file.path("R", "benchmarks.R")) model_names <- list( ar.x="AR", @@ -3628,7 +3628,6 @@ arma.print_openmp_vs_opencl(model_names, row_names) #+caption: Running time (s.) for OpenMP and OpenCL implementations of AR, MA and LH models. #+attr_latex: :booktabs t #+RESULTS: tab-arma-performance -#+BEGIN_SRC org | | | | OpenMP | | OpenCL | | \orgcmidrule{2-4}{5-6}Subroutine | AR | MA | LH | AR | LH | |----------------------------------+------+------+--------+--------+--------| @@ -3636,7 +3635,6 @@ arma.print_openmp_vs_opencl(model_names, row_names) | Validate model | 0.09 | 0.11 | | 0.08 | | | Generate wavy surface | 1.26 | 5.60 | 351.68 | 769.55 | 25.05 | | Compute velocity potentials | 0.02 | 0.01 | 0.02 | 0.01 | 0.01 | -#+END_SRC In contrast to AR model, LH model exhibits the best performance on GPU and the worst performance on CPU. The reasons for that are