Skip to content
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

Jump to bottom

add Python 3.11 correction factor #10

Draft
wants to merge 4 commits into
base: master
Choose a base branch
from
Draft

add Python 3.11 correction factor #10

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9ca46c5
add correction factor for python 3.11
Nov 13, 2023
be4efef
add same factor for AMD and Intel cpus
Nov 15, 2023
e7c20b2
retrieve cpu brand via /proc/cpuinfo/vendor_id instead of /proc/cpuin…
Nov 20, 2023
5a864df
Merge branch 'DIRACGrid:master' into master
EwoudK Nov 23, 2023
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
6 changes: 5 additions & 1 deletion src/db12/benchmark.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -228,7 +228,11 @@ def get_norm_correction(norm_computed):
try:
with open("/proc/cpuinfo", "r") as file_object:
content = file_object.read()
cpu_brand_name = re.findall("model name\t: ([a-zA-Z]*)", content)[0]
cpu_brand_name = re.findall("vendor_id\t: ([a-zA-Z]*)", content)[0]
if cpu_brand_name == "GenuineIntel":
cpu_brand_name = "Intel"
elif cpu_brand_name == "AuthenticAMD":
cpu_brand_name = "AMD"
except (IOError, IndexError):
logging.warning(
"Cannot correct the score, return the raw norm: cannot access CPU information"
Expand Down
6 changes: 6 additions & 0 deletions src/db12/factors.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,12 @@
"Intel": 0.86,
"AMD": 0.71
}
},
"3.11": {
"cpu_brand": {
"Intel": 0.912400392678881,
"AMD": 0.912400392678881
}
Comment on lines +8 to +13
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If I understand correctly, these factors aim at aligning the DB12 score, computed with python3.11, with the HEPScore. Am I right?

factors.json was originally introduced to align the original DB12 score, computed with python2, with the DB12 score computed with python3.9.

The questions are:

  • do we want to introduce this HEPScore scaling factor directly in the DB12 repo?

    • DB12 can also be used outside of the LHC context
    • The HEPSPEC06 scaling factor was set in LHCbDIRAC

  • do we still want to be aligned with the original DB12 score (python2)?

    • DB12 can be used outside of the LHC context again.
    • Regarding LHCb MC sim applications, the raw python3.9 DB12 score was more accurate than the original DB12 score on AMD CPUs (and a bit less accurate on Intel but not significantly...)
    • May be we need to compare the raw python3.11 DB12 score to the LHCb MC sim applications now.

Any opinion about that @fstagni?

}
}
}
Loading