- fisyak's home page
- Posts
- 2024
- November (1)
- October (4)
- September (3)
- August (6)
- July (3)
- June (3)
- March (1)
- February (4)
- January (2)
- 2023
- December (2)
- November (1)
- October (2)
- September (3)
- August (3)
- July (2)
- June (4)
- April (2)
- March (2)
- February (1)
- January (2)
- 2022
- 2021
- December (2)
- October (2)
- September (1)
- August (2)
- June (3)
- May (1)
- April (2)
- March (3)
- February (5)
- January (4)
- 2020
- 2019
- 2018
- 2017
- December (3)
- October (2)
- September (1)
- August (7)
- July (3)
- May (2)
- April (1)
- February (1)
- January (2)
- 2016
- 2015
- November (1)
- October (2)
- September (4)
- August (1)
- July (1)
- June (2)
- May (4)
- April (1)
- March (1)
- February (3)
- January (1)
- 2014
- 2013
- December (4)
- My blog
- Post new blog entry
- All blogs
Compare cpu per event for gcc 4.8.2, 4.9.2, 5.2.1 and icc 17.0.0 compilers for 2013 pp510 W sample on HLT farm
After merging online and offline CA codes and switching to Vc version 1.2
I have made timing measurement of relative perfomance of gcc 4.8.2, 4.9.2, 5.2.1 and icc 17.0.0 compilers
and summarize them in the table below.
gcc 4.8.2 |
gcc 4.9.2 | gcc 5.2.1 | icc 17.0.0 | |||||||||||
-m32 | -m64 | -m32 | -m64 | -m32 |
-m64 |
-m32 |
||||||||
-g | -g -O2 | -g | -g -O2 | -g | -g -O2 | -g | -g -O2 | -g | -g -O2 | -g | -g -O2 | -g | -g -O2 | |
CPU(sec) | 37.70 | 16.61 | 34.18 | 13.18 | 39.30 | 18.46 | 32.91 | 15.10 | 40.16 | 19.02 | 33.12 | 13.85 | 17.84 | 17.89 |
Ratio | 2.27 | 1.00 | 2.06 | 0.79 | 2.37 | 1.11 | 1.98 | 0.91 | 2.42 | 1.14 | 1.99 | 0.83 | 1.07 | 1.08 |
The comments and conclusions from the above table:
1. The default optimization for icc is -O2 (instead -O0 for gcc). This is the reason why -g and -g -O2 give the same result.
2. -m64 option with icc does not work because it requires rebuilding of cernlib.
3. gcc 5.2.1 from the standard Software Collections does not support ia32 architecture (-m32).
4. A CPU gain ~20% has been observed for optimized verion with x8664 (intel64, -m64) over ia32 (-m32) architectures.
5. There no cpu gain with icc with respect to gcc 4.9.2 (-m32) and there is a degradate ~8% with respect to gcc 4.8.2 (-m32).
6. It does not look like that we can gain anything essential from icc.
I think icc thread can be considered as non profitable in the nearest future.
- fisyak's blog
- Login or register to post comments