Skip to content
GitLab
Explore
Sign in
Register
Primary navigation
Search or go to…
Project
F
fluiddyn_papers
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container registry
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Terms and privacy
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
fluiddyn
fluiddyn_papers
Commits
61402ef0
Commit
61402ef0
authored
4 years ago
by
Pierre Augier
Browse files
Options
Downloads
Patches
Plain Diff
Finalize
parent
8a67cdec
No related branches found
No related tags found
1 merge request
!5
First modifications to take care of the editor remarks
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
reply_Zwart2020/correspondence.md
+7
-7
7 additions, 7 deletions
reply_Zwart2020/correspondence.md
reply_Zwart2020/reply_editor_NA.md
+8
-7
8 additions, 7 deletions
reply_Zwart2020/reply_editor_NA.md
with
15 additions
and
14 deletions
reply_Zwart2020/correspondence.md
+
7
−
7
View file @
61402ef0
...
...
@@ -137,7 +137,7 @@
implementations using Pythran+OpenMP and Julia executed using 6 and 12 CPU
cores. We consider in Fig.
\r
ef{fig1} the energy consumption of the cores used
(6 or 12 for these runs and 1 for the sequential jobs), which make sense on
shared cluster
node
s wherein on
ly requested cores are reserved
. We see that
shared clusters wherein on
e can reserve only the needed cores
. We see that
parallelism with threads has only a moderate impact on energy consumption since
the increase in power consumption partly counterbalances the decrease of
elapsed time. For example, the 12-threaded Pythran version is 10 times faster
...
...
@@ -149,12 +149,12 @@
Python can actually be good solutions to easily obtain good performance while
retaining simplicity and readability. We think that minimizing the ecological
impact of scientific computing is mainly limited by human factors: time, work,
knowledge and skills. For example, scientists have to be able to
run heavy
computations on clusters with workload manag
ers optimized in terms of energy
consumption. They
should know how to profile their code to discover which parts
can potentially
be optimized. Therefore, time and money should be invested in
education and
tooling to minimize the overall ecological impact of computing,
irrespective of
the underlying language.
knowledge and skills. For example, scientists have to be able to
use job
schedulers and shared clust
ers optimized in terms of energy
consumption. They
should know how to profile their code to discover which parts
can potentially
be optimized. Therefore, time and money should be invested in
education and
tooling to minimize the overall ecological impact of computing,
irrespective of
the underlying language.
\s
ubsection
*
{Acknowledgements}
...
...
This diff is collapsed.
Click to expand it.
reply_Zwart2020/reply_editor_NA.md
+
8
−
7
View file @
61402ef0
# What do we reply
Thank you for reading our draft and for your suggestions.
Thank you for reading our draft and for your suggestions. The pdf of a new
version is attached.
We agree with your suggestion to focus more on the commonalities between our
position and that of Zwart. We tried to be simple and short but we left the
...
...
@@ -12,8 +13,8 @@
language on the ecology" and in the conclusions. We rewrote the first paragraph
which is now:
"We read with attention the comment by Zwart on
**
The ecological impact of
high-performance computing in astrophysics
**
\c
ite{Zwart2020}. We fully agree
"We read with attention the comment by Zwart on The ecological impact of
high-performance computing in astrophysics
\c
ite{Zwart2020}. We fully agree
with its take-home message: scientists should be mindful of their carbon
footprint. One of the proposed solutions is to avoid the Python programming
language. We advocate that this would be counterproductive and that scientific
...
...
@@ -39,7 +40,7 @@
code is to interpret it with the reference implementation called CPython. Due
to the lack of a built-in JIT compiler, CPython is relatively slow."
In total, with the slightly longer opening paragraph and no footnotes, the
full
text
is approximately 140 words shorter. We also took into account your other
pieces
of advice. We hope that this manuscript can be suitable for publication
in this
state.
In total, with the slightly longer opening paragraph and no footnotes, the
text
is approximately 140 words shorter. We also took into account your other
pieces
of advice. We hope that this manuscript can be suitable for publication
in this
state.
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment