TELCODOCS-2057: First draft

_topic_maps/_topic_map.yml

@@ -3560,6 +3560,8 @@ Topics:
   File: nvidia-gpu-architecture
 - Name: AMD GPU Operator
   File: amd-gpu-operator
+- Name: Intel Gaudi AI accelerators
+  File: gaudi-ai-accelerator
 ---
 Name: Backup and restore
 Dir: backup_and_restore

hardware_accelerators/gaudi-ai-accelerator.adoc

@@ -0,0 +1,22 @@
+:_mod-docs-content-type: ASSEMBLY
+[id="gaudi-ai-accelerator"]
+= Intel Gaudi AI accelerators
+include::_attributes/common-attributes.adoc[]
+:context: gaudi-ai-accelerator
+
+toc::[]
+
+:FeatureName: Intel Gaudi AI accelerators
+
+You can use Intel Gaudi AI accelerators for your {product-title} generative AI and machine learning (AI/ML) applications. Intel Gaudi AI accelerators offer a cost-efficient, flexible, and scalable solution for optimized deep learning workloads.
+
+Red{nbsp}Hat supports Intel Gaudi 2 and Intel Gaudi 3 devices. Intel Gaudi 3 devices provide significant improvements in training speed and energy efficiency.
+
+include::modules/gaudi-ai-accelerator-prerequisites.adoc[leveloffset=+1]
+
+[role="_additional-resources"]
+.Additional resources
+
+* link:https://docs.habana.ai/en/latest/Installation_Guide/Additional_Installation/OpenShift_Installation/index.html[OpenShift Installation] (Intel Gaudi documentation)
+
+* link:https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/2.20/html/working_with_accelerators/intel-gaudi-ai-accelerator-integration_accelerators#intel-gaudi-ai-accelerator-integration_accelerators[Intel Gaudi AI Accelerator integration]

modules/gaudi-ai-accelerator-prerequisites.adoc

@@ -0,0 +1,16 @@
+// Module included in the following assemblies:
+//
+// * hardware_accelerators/gaudi-ai-accelerator.adoc
+
+
+:_mod-docs-content-type: CONCEPT
+[id="gaudi-ai-accelerators-prerequisites_{context}"]
+= Intel Gaudi AI accelerators prerequisites
+
+* You have a working {product-title} cluster with at least one GPU worker node.
+
+* You have access to the {product-title} cluster as a cluster-admin to perform the required steps.
+
+* You have installed {oc-first}.
+
+* You have installed the Node Feature Discovery (NFD) Operator and created a `NodeFeatureDiscovery` instance.