36
edits
Changes
mLine 3:
Line 3: − + − + − + − + − + + +
no edit summary
<code>
<code>
<pre>
<pre>
./gnarly.py n | time java -Xmx1024M Main > /dev/null
./gnarly.py n | time java -Xmx8192M -Xincgc Main > /dev/null
</pre>
</pre>
</code>
</code>
where n is replaced with one of the values from the table. (You may need to <code>tap java150</code> to access Java on Grace.)
where n is replaced with one of the values from the table. (You may need to <code>tap java150</code> to access Java on Grace.) Yes, that's a lotta memory.
{| border="1" cellpadding="6" cellspacing="0" align="center" style="margin: 1em auto 1em auto; width:50%"
{| border="1" cellpadding="6" cellspacing="0" align="center" style="margin: 1em auto 1em auto"
|-
|-
!Who !! n=10 !! n=100 !! n=1000 !! n=10000
!Who !! n=10 !! n=100 !! n=1000 !! n=5000 !! Notes
|-
|-
|Ben || 1 || 1 || 11 || Still running...
|Ben || 1 || 1 || 11 || 210 || Run by jkleint... Used 7 GB RAM. That's 1.4 MB per city. Dang.
|}
|}
This is just for kicks on Part 1; we're not grading performance. These timings probably include so much XML parsing overhead as to be pretty meaningless anyway. They're a rough guide; if your solution is working and you're taking more than 2-3x this time (or memory!), try running a profiler to see where your code is spending its time. Or steal someone else's for Part 2. :)