diff --git a/qa/cft/FeatureBaseClusterCFTTremor.yaml b/qa/cft/FeatureBaseClusterCFTTremor.yaml
index 3cd001869..7820148c5 100644
--- a/qa/cft/FeatureBaseClusterCFTTremor.yaml
+++ b/qa/cft/FeatureBaseClusterCFTTremor.yaml
@@ -44,7 +44,7 @@ Parameters:
     Description: The number of EC2 instances for the FeatureBase cluster  
   FBInstanceSize:
     Type: String
-    Default: m5ad.16xlarge
+    Default: m6i.16xlarge
     Description: Class of EC2 instance used to host the deployment.  
   # note that we are using a producer, which is also the kafka host
   # and the consumer uses the kafka-static binary for ingesting data
@@ -52,10 +52,10 @@ Parameters:
   # producer/consumer related params  
   ProducerInstanceSize:
     Type: String
-    Default: m4.16xlarge
+    Default: m6i.16xlarge
   ConsumerInstanceSize:
     Type: String
-    Default: m4.16xlarge
+    Default: m6i.16xlarge
   ProducerInstanceCount:
     Type: Number
     Default: 1
diff --git a/qa/tf/.modules/featurebase-cluster/variables.tf b/qa/tf/.modules/featurebase-cluster/variables.tf
index 5e9d59dd4..5a743f703 100644
--- a/qa/tf/.modules/featurebase-cluster/variables.tf
+++ b/qa/tf/.modules/featurebase-cluster/variables.tf
@@ -10,7 +10,7 @@ variable "fb_cluster_arch" {
 
 variable "fb_ingest_type" {
   type    = string
-  default = "c6g.2xlarge"
+  default = "m6g.2xlarge"
 }
 
 variable "fb_ingest_node_count" {
@@ -20,7 +20,7 @@ variable "fb_ingest_node_count" {
 
 variable "fb_data_node_type" {
   type    = string
-  default = "c6g.16xlarge"
+  default = "m6g.16xlarge"
 }
 
 variable "fb_data_node_count" {
diff --git a/qa/tf/ci/deletetest/delete_cloud_init.sh b/qa/tf/ci/deletetest/delete_cloud_init.sh
index 6c4c48a0d..34ca89135 100644
--- a/qa/tf/ci/deletetest/delete_cloud_init.sh
+++ b/qa/tf/ci/deletetest/delete_cloud_init.sh
@@ -7,6 +7,7 @@ echo "Running delete_cloud_init.sh"
 
 # create and initialize a raid0 array 
 # for m5ad.16xlarge instance type, there are 4 NVMe SSDs. 
+# Note JALLISON 20221214 - we are now using m6i instance types for cost savings
 sudo mdadm --create --verbose /dev/md0 --level=0 --raid-devices=4 /dev/nvme1n1 /dev/nvme2n1 /dev/nvme3n1 /dev/nvme4n1
 
 # format filesystem 
diff --git a/qa/tf/ci/deletetest/main.tf b/qa/tf/ci/deletetest/main.tf
index 3d6bd2c77..aa1188822 100644
--- a/qa/tf/ci/deletetest/main.tf
+++ b/qa/tf/ci/deletetest/main.tf
@@ -4,10 +4,10 @@ module "ci-cluster" {
     region = var.region
     profile = var.profile
     fb_cluster_arch = ["x86_64"]
-    fb_data_node_type = "m5ad.16xlarge" 
+    fb_data_node_type = "m6i.16xlarge" 
     fb_data_node_count = 5
     ebs_volumes = []
-    fb_ingest_type = "m4.16xlarge"
+    fb_ingest_type = "m6i.16xlarge"
     fb_ingest_node_count = 2
     fb_ingest_disk_size_gb = 1200
     vpc_id = "vpc-05a26a122f961dc2b"
diff --git a/qa/tf/perf/able/main.tf b/qa/tf/perf/able/main.tf
index 508153d21..10a9054b7 100644
--- a/qa/tf/perf/able/main.tf
+++ b/qa/tf/perf/able/main.tf
@@ -3,10 +3,10 @@ module "able-cluster" {
     cluster_prefix = var.cluster_prefix
     region = var.region
     profile = var.profile
-    fb_data_node_type = "c6g.16xlarge"
+    fb_data_node_type = "m6g.16xlarge"
     fb_data_disk_iops = 10000
     fb_data_node_count = 3
-    fb_ingest_type = "c6g.2xlarge"
+    fb_ingest_type = "m6g.2xlarge"
     fb_ingest_disk_iops = 10000
     fb_ingest_disk_size_gb = 500
     fb_ingest_node_count = 1