Configuring GPU Workloads for Raster Analytics on ArcGIS Enterprise on Kubernetes

Question

Any success stories or tips for running GPU-backed raster analysis workloads (deep learning, image classification) on ArcGIS Enterprise on Kubernetes? Looking to optimize tile caching and GPU scheduling across nodes.

VenkataKondepati · Answer

Hi @Priya_Das,

Thanks for the details. Sure, I will take a look and let you know.

Regards,
Venkat

Priya_Das · Answer

Hi @VenkataKondepati,

You can refer to following documentation:

* https://enterprise-k8s.arcgis.com/en/11.4/deploy/gpu-enabled-nodes.htm
* https://enterprise-k8s.arcgis.com/en/11.4/administer/configure-raster-analytics.htm
* https://enterprise-k8s.arcgis.com/en/latest/analyze/deep-learning-in-raster-analysis.htm