https://community.esri.com/t5/arcgis-api-for-python-questions/rasterprocessinggpu-service-cannot-release-gpu/m-p/1591565#M11195 <P>l used arcgis python api to perform deeplearning inference tasks. When one task successed, proceed to another task, the job of&nbsp;RasterAnalysisTools failed. Job messages prompted GPU memory is occupied.</P><P>l found instance of RasterProcessingGPU&nbsp; do not release GPU memory.</P><P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="FrankLee1_0-1741073988105.png" style="width: 400px;"><img src="https://community.esri.com/t5/image/serverpage/image-id/126829iBEF80B264F14335A/image-size/medium?v=v2&amp;px=400" role="button" title="FrankLee1_0-1741073988105.png" alt="FrankLee1_0-1741073988105.png" /></span></P><P>l tried some ways but not&nbsp;optimal.</P><P>1. Adjust RasterprocessingGPU service recycling interval to short time</P><P>2 Restarted&nbsp;&nbsp;RasterprocessingGPU service after&nbsp;each execution of the script.</P><P>But my question is&nbsp; instance of service can release GPU memory automatically ?</P> Wed, 05 Mar 2025 08:40:15 GMT FrankLee1 2025-03-05T08:40:15Z https://community.esri.com/t5/arcgis-api-for-python-questions/rasterprocessinggpu-service-cannot-release-gpu/m-p/1591565#M11195 <P>l used arcgis python api to perform deeplearning inference tasks. When one task successed, proceed to another task, the job of&nbsp;RasterAnalysisTools failed. Job messages prompted GPU memory is occupied.</P><P>l found instance of RasterProcessingGPU&nbsp; do not release GPU memory.</P><P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="FrankLee1_0-1741073988105.png" style="width: 400px;"><img src="https://community.esri.com/t5/image/serverpage/image-id/126829iBEF80B264F14335A/image-size/medium?v=v2&amp;px=400" role="button" title="FrankLee1_0-1741073988105.png" alt="FrankLee1_0-1741073988105.png" /></span></P><P>l tried some ways but not&nbsp;optimal.</P><P>1. Adjust RasterprocessingGPU service recycling interval to short time</P><P>2 Restarted&nbsp;&nbsp;RasterprocessingGPU service after&nbsp;each execution of the script.</P><P>But my question is&nbsp; instance of service can release GPU memory automatically ?</P> Wed, 05 Mar 2025 08:40:15 GMT https://community.esri.com/t5/arcgis-api-for-python-questions/rasterprocessinggpu-service-cannot-release-gpu/m-p/1591565#M11195 FrankLee1 2025-03-05T08:40:15Z