Adding GKE CPU Direct TCPX example

Author: barney-s

examples/gcp/gke-gpudirect-tcpx-a3mega/README.md

@@ -0,0 +1,178 @@
+# GkeGpuDirectTCPXCluster
+
+This example deploys [Google Cloud GPU supercomputer](https://cloud.google.com/kubernetes-engine/docs/how-to/gpu-bandwidth-gpudirect-tcpx) which is accelerator-optimized for scalable, massive models. The RGD is installed by platform administrators who facilitate ML infrastructure for self service teams.
+
+The cluster has:
+* Eight NVIDIA H100 GPUs per machine.
+* Up to 200 Gbps bandwidth on the primary NIC.
+* Secondary NICs (up to four on A3 High machine types), each supporting up to 200 Gbps bandwidth for GPU data transfer.
+
+This deployment maximizes network bandwidth and throughput for high-performance GPU workloads in Google Kubernetes Engine (GKE) Standard clusters by using, GPUDirect-TCPX, gVNIC, and multi-networking. 
+
+* GKE cluster
+* Container Node Pools
+* Network
+* Subnetwork
+* GKE Network and NetworkParams
+
+Everything related to these resources would be hidden from the end user, simplifying their experience.  
+
+![GKE GPU A3Mega](gke-gpudirect-a3mega.png)
+
+<!--
+meta {
+  title "Gke GpuDirect TCPX Cluster"
+}
+
+elements {
+  gcp {
+      group k8sconfig {
+      name "Kubernetes Manifests"
+      card kubernetes as config1 {
+         name "Network"
+      }
+      card kubernetes as config2 {
+         name "GKENetworkParamSet "
+      }
+    } 
+
+    group Network {
+      card firewall as fw1 {
+         name "firewall 1"
+      }
+      card firewall as fw2 {
+         name "firewall 2"
+      }
+      card firewall as fw3 {
+         name "firewall 3"
+      }
+      card firewall as fw4 {
+         name "firewall 4"
+      }
+  
+      card network as net1 {
+         name "net 1"
+      }
+      card network as net2 {
+         name "net 2"
+      }
+      card network as net3 {
+         name "net 3"
+      }
+      card network as net4 {
+         name "net 4"
+      }
+      card network as snet1 {
+         name "subnet 1"
+      }
+      card network as snet2 {
+         name "subnet 2"
+      }
+      card network as snet3 {
+         name "subnet 3"
+      }
+      card network as snet4 {
+         name "subnet 4"
+      }
+    }
+    group GKE {
+      card gke as cluster {
+         name "cluster"
+      }
+      
+      group default {
+        name "Default Nodepool"
+        card gke as defaultNodepool {
+         name "nodepool"
+        }
+        card gce as generalVM {
+         name "e2-medium"
+        }
+      }
+      
+      group gpu {
+        name "GPU Nodepool"
+        card gke as gpuNodepool {
+         name "nodepool "
+        }
+        card gce as gpuVM {
+         name "a3-highgpu-8g"
+        }
+        card gpu as nvidia {
+         name "Nvidia H100"
+        }
+      }
+      
+    }
+   
+ }
+}
+
+paths {
+  fw1 -\-> net1
+  fw2 -\-> net2
+  fw3 -\-> net3
+  fw4 -\-> net4
+  
+  net1 -\-> snet1
+  net2 -\-> snet2
+  net3 -\-> snet3
+  net4 -\-> snet4
+  
+  config1 -\-> config2
+  
+  defaultNodepool -\-> generalVM
+  gpuNodepool -\-> gpuVM
+  gpuVM -\-> nvidia
+
+  Network -right-> GKE
+  k8sconfig -right-> cluster
+}
+-->
+
+
+## End User: GkeGpuDirectTCPXCluster
+
+The administrator needs to install the RGD first.
+The end user creates a `GkeGpuDirectTCPXCluster` resource something like this:
+
+```yaml
+apiVersion: kro.run/v1alpha1
+kind: GkeGpuDirectTCPXCluster
+metadata:
+  name: gpu-demo
+  namespace: config-connector
+spec:
+  name: gpu-demo        # Name used for all resources created as part of this RGD
+  location: us-central1 # Region where the GCP resources are created
+```
+
+They can then check the status of the applied resource:
+
+```
+kubectl get gkegpudirecttcpxcluster
+kubectl get gkegpudirecttcpxcluster gpu-demo -n config-connector -o yaml
+```
+
+Navigate to GKE Cluster page in the GCP Console and verify the cluster creation.
+
+Once done, the user can delete the `GkeGpuDirectTCPXCluster` instance:
+
+```
+kubectl delete gkegpudirecttcpxcluster gpu-demo -n config-connector
+```
+
+## Administrator: ResourceGraphDefinition
+The administrator needs to install the RGD in the cluster first before the user can consume it:
+
+```
+kubectl apply -f rgd.yaml
+```
+
+Validate the RGD is installed correctly:
+
+```
+kubectl get rgd gkegpudirecttcpxcluster.kro.run
+```
+
+Once all user created instances are deleted, the administrator can choose to deleted the RGD.