[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[gnugo-devel] update to break-in timings comparison
From: |
Arend Bayer |
Subject: |
[gnugo-devel] update to break-in timings comparison |
Date: |
Tue, 24 Jun 2003 12:40:44 +0200 (CEST) |
In summary, the break-in code costs about 10% on average, and 20% in
worst case. It's still not as fast as I would like it to be, but by now
my impression is that it is worth it. Opinions appreciated.
We are a lot faster than 3.2. We could really increase some depth levels
for 3.4 if only we knew which ones are worth it.
Arend
(The comparison between 3.3.21, no breakin and cvs is not 100% fair as
paul_3_22.2a is a small speed-up. So add 1% or so to the cvs figures.)
File:
3.3.21, no breakin 3.3.21 with breakin cvs with patch 3.2
gnugo-3.3.15-Kwaku-200301281930.gtp.in:
1086 1144 1118 1065
+5.3% +2.9%
gnugo-3.3.15-morli-200301291908.gtp.in:
1057 1214 1149 2199
+14.8% +8.7%
gnugo-3.3.15-wingjk-200301260551.gtp.in:
1104 1439 1246 1901
+30.3% +12.8%
gnugo-go4.gtp.in:
1544 2221 1855 1529
+43.8% +20.1%
kit-gnugo-3.3.15-200301282141.gtp.in:
1114 1425 1229 1107
+27.9% +10.3%
leftd-gnugo-3.3.15-200301260333.gtp.in:
1683 1748 1688 2678
+3.8% +0.2%
manyfaces-gnugo.gtp.in:
999 1148 1123 1622
+14.9% +12.4%
Total:
8587 10342 9408 12101
+20.4% +9.5%
- [gnugo-devel] update to break-in timings comparison,
Arend Bayer <=