Received: by ATHENA-PO-2.MIT.EDU (5.45/4.7) id AA22131; Fri, 9 Jun 89 19:14:03 EDT
From: <russ@ATHENA.MIT.EDU>
Received: by ATHENA.MIT.EDU (5.45/4.7) id AA07513; Fri, 9 Jun 89 19:15:32 EDT
Received: by HASTUR.MIT.EDU (5.45/4.7) id AA22187; Fri, 9 Jun 89 19:13:31 EDT
Date: Fri, 9 Jun 89 19:13:31 EDT
Message-Id: <8906092313.AA22187@HASTUR.MIT.EDU>
To: bgardner@ATHENA.MIT.EDU
Subject: timing on Xt



I've forgotten the formula for standard deviation etc. (something
about summing the squares of the error) but here's the raw data.

Again, the sources and the test run is in:

	/mit/visual2/russ/testXt

Not sure why I had a couple of extra-long waits in the 
"pre,1x1" case.

--Russ



==========
post,1x1:

2400
2170
2150
1900
1910
1840
1850
1920
1850
1840

=================
post,batch:

1980
2080
2200
2170
2000
2000
2230
2080
2000
2310

============
pre,1x1:

10920
4650
4590
4620
4700
4600
11010
4630
4610
4600

=====================
pre,batch:

2890
2820
2910
2880
2880
2840
2740
2880
2850
2870