<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;"><br>On Mar 23, 2014, at 12:29 PM, Jed Brown <<a href="mailto:jed@jedbrown.org">jed@jedbrown.org</a>> wrote:<br><br><blockquote type="cite">Handy visualization, though I think the red text is excessive [1]</blockquote><div><br></div> Ok, I’ve removed it.<div><br><blockquote type="cite"> and as<br>something unrelated from correctness of the installation, I'm not sure<br>it should be displayed as part of "make all", especially when it doesn't<br>run on batch systems.<br></blockquote><div><br></div> In theory it only prints this suggestion on machines without a batch system and it is printed after make test. Not make all.</div><div><br></div><div>test:<br><span class="Apple-tab-span" style="white-space:pre"> </span>-@${OMAKE} PETSC_ARCH=${PETSC_ARCH} PETSC_DIR=${PETSC_DIR} test_build 2>&1 | tee ./${PETSC_ARCH}/conf/test.log<br><span class="Apple-tab-span" style="white-space:pre"> </span>-@if [ "${PETSC_WITH_BATCH}" == "" ]; then printf "Now run make streams NPMAX=<number of expect MPI processes you intend to use>\nto evaluate the computer systems you plan use\n”;fi</div><div><br></div><div> Is this not working and it is always printing the suggestion?</div><div><br><blockquote type="cite"><br>I see a lot of variability between runs (e.g., erratic performance on<br><a href="http://es.mcs.anl.gov">es.mcs.anl.gov</a> attached), so perhaps we should be making box-and-whisker<br>plots. This is a balancing act because we don't want the test to take a<br>long time to run.<br><br>While the "ideal speedup" line is visceral, it makes it hard to see<br>what's actually happening. I would either remove that line or add a<br>second y axis that is unscaled (second attachment). Either way, we<br>should find a way to report numbers rather than unscaled speedup.<br></blockquote><div><br></div> In my latest push, from this morning, I added the bandwidths on the right axis.</div><div><br></div><div> I agree with Matt I don’t like the other plot overlaid on top. Too difficult to understand. I also don’t like a separate plot, nor do I want to remove the perfect scaling because on a decent system it provides useful information. I have tried the following: the top of the prefect scaling will be removed to make sure that at least 1/3 of the y axis is used to display actual scaling. See attached image (of course it could be changed to use 1/2 the y axis or whatever.) Is this a reasonable compromise.<br><blockquote type="cite"><br>It's crazy that a plot of intra-node scalability of STREAMS is not<br>posted prominently in computing facility documentation.<br></blockquote><div><br></div> It is disgraceful but not surprising.<img height="432" width="576" apple-width="yes" apple-height="yes" apple-inline="yes" id="11D0F920-A5BD-445D-95F3-FEDBB21ABC6F" src="cid:FFC86997-70F5-4ED7-BA38-42FACCA6D4F0@wowway.com"><br><blockquote type="cite"><br><br>[1] I still disagree with using colored text unconditionally. I don't<br>want control characters in log files and emails. We could test whether<br>output is going to a TTY so long as a separate stream is being sent to<br>the screen than the log files.<br><br><br><span><scaling-es.png></span><span><scaling.png></span><br></blockquote><br></div></body></html>