Differences

This shows you the differences between two versions of the page.

--- io500:start [2018/05/08 01:08] – external edit 127.0.0.1
+++ io500:start [2018/11/23 12:10] – kunkel
@@ Line 1: / Line 1: @@
 {{indexmenu_n>2}}
-{{:io500:logo-io500.png?200}}
-<io500>
+{{ :io500:logo-io500.png?200}}
-date=2017-11
+===== IO-500 =====
-</io500>
-===== IO500 List =====
+This is the official IO-500 **ranked list**((That means only those entries that qualify according to the wish of the submitter are contained.)) from November 2018 (from [[https://sc18.supercomputing.org|SC 2018]]).
-The above table shows the current IO500 list as of SC17.  By default the merely shows the top-level summary information.
+Please see also [[io500:list:18-11:10node|the 10 node challenge ranked list]] from November 2018.
-===== Expanding the table =====
+The list shows the best result for a given combination of system/institution/filesystem.
-To show more data, such as the individual benchmark scores, use the "Add column" drop-down box above.  There are shortcuts, such as ior.*, to add all of the individual benchmarks within a group.  The three groups are the bandwidth scores, as measured with ior, the metadata scores, as measured with mdtest, and the find scores (currently just one), as measured with the find utility.
+<io500>
+date=2018-11
-===== Adding custom columns =====
+version=v3
+radar=false
-To add a custom (i.e. derived) column, merely enter an equation in the Equation field above.
+bestOnly=true
+</io500>
-For equations you may use variables, i.e., use the group name of the metrics you want to use and replace space in the variable name with underscore, for example:
-  * io500.score / information.client_nodes
-  * io500.bw / information.client_nodes
-  * mdtest.easy_create / information.client_nodes
-If the metrics name is unambiguous, you may omit the group name, i.e., the following is valid:
-  * score / client_nodes
-  * bw / client_nodes
-  * easy_create / client_nodes
-====== Scoring information ======
-The score by which the ranking is computed is the 'score' column.  This 'score' column is merely the square root of the product of the 'bw' and 'md' scores.  The 'bw' score is the geometric mean of the IOR scores and the 'md' score is the geometric mean of the mdtest scores and the find scores.
+{{https://www.vi4io.org/assets/io500/2018-11/data.csv|Download the full list with all data as CSV}}
-====== Further information ======
-  * [[events:2017:bof-sc-bof|Detailed analysis of the first list are available in the slide deck from our SC BoF]]
+//Values with* indicate that a value for the computation was missing.//
-  * [[std:io500:start|For details of the approach, see the IO-500 effort]]
-====== Submit for the next list ======
+See also [[io500:list:18-11:10node|the 10 node challenge ranked list]].
-Please consider submitting your own result to make it onto the next list!  Fetching, compiling, and running the IO500 Benchmarks is surprisingly straightforward.  Although tuning can be challenging.  Instructions for the former can be found in our [[https://github.com/VI4IO/io-500-dev|github repo]].  Help with the latter can be found in our various [[https://www.vi4io.org/std/io500/start#communication_contribution|communication channels]].
+Further lists with more detail can be found on the navigation menu.
+For example, the full list with [[:io500:list:18-11:full|radar chart and configurable scoring]].

Virtual Institute for I/O

User Tools

Site Tools

Differences

Virtual Institute
for
I/O