# CMAQ benchmark differences **URL:** \ **Category:** Compiling\ **Created:** [January 10, 2019, 10:28pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475 "2019-01-10T22:28:46Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![jdeast2](https://avatars.discourse-cdn.com/v4/letter/j/6a8cbe/32.png) [@jdeast2](https://forum.cmascenter.org/u/jdeast2)\ **Post date:** [January 10, 2019, 10:28pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/1 "2019-01-10T22:28:46Z")
Hello CMAQ users, I’ve compiled CMAQ 5.0.2 using intel 2017. After compiling, the benchmark was successful for most variables, but not all. In general, for particulate & aerosol species it was successful, and for gaseous species it was not successful. I’ve attached some files showing the errors in O3 ASO4. Green area meets the “0.5% error” criteria and any other color fails to meet that criteria. I can provide the config file as well. Has anyone encountered similar error when benchmarking? If so, how did you deal with it? Does anyone with know-how regarding compiling and benchmarking have suggestions for where this error might come from? We tried several flag variations unsuccessfully. Thank you for any help or guidance you can provide! Best, James ![O3diff_animate2](https://canada1.discourse-cdn.com/flex027/uploads/cmas/original/1X/0594ab7c2d07bad729d5826c84f538922d91b384.gif) ![ASO4Jdiff_animate2](https://canada1.discourse-cdn.com/flex027/uploads/cmas/original/1X/91ed2114b41a56361cf258a6661eee4ca03a82f9.gif) ---
**Author:** ![Ben\_Murphy](https://avatars.discourse-cdn.com/v4/letter/b/74df32/32.png) [@Ben\_Murphy](https://forum.cmascenter.org/u/Ben_Murphy)\ **Post date:** [January 15, 2019, 5:33pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/2 "2019-01-15T17:33:53Z")
Hi James, This advice is almost anecdotal but if possible, you can try to update all the way to intel 18. Some of us have noticed some minor difference between versions 18 and 17 for the new CMAQv5.3 but I’m not sure what it would do for version 5.0.2. The spatial extent of the differences look like changes have manifest in inline point source emissions. Cheers, Ben ---
**Author:** ![jdeast2](https://avatars.discourse-cdn.com/v4/letter/j/6a8cbe/32.png) [@jdeast2](https://forum.cmascenter.org/u/jdeast2)\ **Post date:** [January 28, 2019, 5:03pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/3 "2019-01-28T17:03:25Z")
Hi Ben, Thanks for your suggestion. I recompiled with intel 2018 and the model output matches the benchmark well. The input data used was the same in both compilations. Thanks, James ---
**Author:** ![esm](https://avatars.discourse-cdn.com/v4/letter/e/f08c70/32.png) [@esm](https://forum.cmascenter.org/u/esm)\ **Post date:** [August 7, 2019, 1:49am UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/4 "2019-08-07T01:49:25Z")
Hi I benchmarked CCTM, but I don’t know how to calculate the differences between my result files and the output files. Could you guide me, please? Thank you ---
**Author:** ![jdeast2](https://avatars.discourse-cdn.com/v4/letter/j/6a8cbe/32.png) [@jdeast2](https://forum.cmascenter.org/u/jdeast2)\ **Post date:** [August 7, 2019, 2:45am UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/5 "2019-08-07T02:45:44Z")
Hello, I’ve found Verdi to be a useful tool to check the benchmark. In this case, I created a map of the percent difference between the benchmark and my compiled version for several variables. Verdi is available through the CMAS center. The tool nc-compare or “nccmp” at the command line can also be downloaded and installed. It does not produce figures as Verdi does. Best, James ---
**Author:** ![esm](https://avatars.discourse-cdn.com/v4/letter/e/f08c70/32.png) [@esm](https://forum.cmascenter.org/u/esm)\ **Post date:** [August 7, 2019, 2:01pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/6 "2019-08-07T14:01:56Z")
Hi Thank you so much. I try to install VERDI, but after installing, I can’t open any files. Could you help me and tell me which version of VERDI do you use and how do you install it? ---
**Author:** ![esm](https://avatars.discourse-cdn.com/v4/letter/e/f08c70/32.png) [@esm](https://forum.cmascenter.org/u/esm)\ **Post date:** [August 8, 2019, 10:13am UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/7 "2019-08-08T10:13:23Z")
Hi Why we have for every output 2 files in output files and 1 file in our result? which of them is right? ---
**Author:** ![jdeast2](https://avatars.discourse-cdn.com/v4/letter/j/6a8cbe/32.png) [@jdeast2](https://forum.cmascenter.org/u/jdeast2)\ **Post date:** [August 8, 2019, 12:50pm UTC](https://forum.cmascenter.org/t/cmaq-benchmark-differences/475/8 "2019-08-08T12:50:02Z")
Hello, I have version 1.5 installed. I followed the install instructions included with the download (for example, INSTALL.linux.txt for Linux systems). You may want to try a different version if your current version doesn’t work. Best, James