upd

teuben · Nov 20, 2023 · 13f08fe · 13f08fe
1 parent a13247d
commit 13f08fe
Showing 1 changed file with 20 additions and 8 deletions.
diff --git a/man/man5/bench.5 b/man/man5/bench.5
@@ -1,4 +1,4 @@
-.TH BENCH 5NEMO "31 August 2023"
+.TH BENCH 5NEMO "20 November 2023"
 
 .SH "NAME"
 bench \- various NEMO benchmarks
@@ -14,19 +14,29 @@ nemobench COMMAND
 There are several methods to compare the performance of NEMO on different processors or environments:
 
 .RS 2
+
+.IP 0.
+The command \fBmake bench5\fP from the $NEMO directory runs a small number of benchmarks designed
+to each takae 5 seconds on a given CPU. An average \fIspeed\fP is then computed to compare different
+compilers and/or CPU's, an idea simular to the \fIgeekbench\fP suite of tests.
+
 .IP 1.
 The command \fBmake bench\fP from the $NEMO directory runs a standard benchmark of about a dozen 
 selected NEMO programs.  It should run in about 30 seconds on a ~2020 computer. A short
 table (see below in RESULTS) lists results from some recent computers.
 
 .IP 2.
 Various \fBBenchfile\fP filles scattered through the $NEMO source tree contain benchmarks. There is
-no  wrapper like "make check" has for Testfile's
+no wrapper like "make check" has for Testfile's
 
 .IP 3.
 Selected man pages have some examples. Over time these should be moved into the Benchfile in
 the source directory where this program is maintained. Some are mentioned below in this man page.
 
+.RS -2
+.PP
+Unless otherwise noted, benchmarks are always performed at the max power settings of a laptop.
+
 .SH "NEMOBENCH5"
 
 The NEMOBENCH5 benchmark (cd $NEMO;make bench5) runs a number (currently 6) N-body codes all calibrated to
@@ -60,16 +70,18 @@ i5-10400H		700	Dell Precision 3351 (astroumd)
 AMD EPYC 7302 @ 3.0GHz	630	lma 1000/ 28000
 Intel Core i9-9900X	-	geminid 1214 / 9467
 Xeon Gold 5218 @ 2.30 GHz	761	unity/node99 1100 / 8610
-i5-10210U CPU @ 1.60 GHz	732 	X1Y4  1029 / 3194
+i5-10210U @ 1.60 GHz	732 	X1Y4  1029 / 3194
 Xeon(R) Silver 4114 @ 2.20GHz	-	terra 777 / 10213
-i7-8550U CPU @ 1.80 GHz	548 	T480 (1600) 696 / 1935
+i7-8550U @ 1.80 GHz	548 	T480 (1600) 696 / 1935
 Xeon E-2186G 3.80 GHz	- 	aurora 1313 / 6721
 i7-3820QM @ 2.70 GHz	498	MacBookPro (Retina, Mid 2012)
-i7-3820 CPU @ 3.60 GHz	414 	dante 833 / 3693
-Xeon(R) CPU E3-1280 @ 3.50 GHz	384	chara 828 / 3117
-i7-3630QM CPU @ 2.40 GHz	429	T530 (1200) 762 / 3006 
+i7-3820 @ 3.60 GHz	414 	dante 833 / 3693
+Xeon(R) E3-1280 @ 3.50 GHz	384	chara 828 / 3117
+i7-3630QM @ 2.40 GHz	429	T530 (1200) 762 / 3006 
 Xeon E5-2623 v4 @ 2.60 GHz	-	kraken 782 / 5800
 Xeon X5550  @ 2.67 GHz	256	sdp 566 / 3938
+AMD Opteron(tm) 6380	234	yorp (2023)
+
 .fi
 
 A typical output of bench5 might look as follows:
@@ -164,7 +176,7 @@ The \fBbench8\fP measures how well your CPU performs simple OpenMP algorithm
 as more cores are employed. This can be an effective way to determine how
 well your CPU adjusts under increased power demands, for example laptops with
 thermal protection will scale down their cpu frequency as more cores are
-employed.
+employed or as a long benchmark heats up the CPU.
 
 .nf
    scaling2 umax=20000 np=1 iter=20