diff --git a/test_common/harness/errorHelpers.cpp b/test_common/harness/errorHelpers.cpp
index c2e1e5256b..f9ca3d6092 100644
--- a/test_common/harness/errorHelpers.cpp
+++ b/test_common/harness/errorHelpers.cpp
@@ -190,6 +190,7 @@ const char *GetChannelTypeName(cl_channel_type type)
         case CL_UNORM_SHORT_565: return "CL_UNORM_SHORT_565";
         case CL_UNORM_SHORT_555: return "CL_UNORM_SHORT_555";
         case CL_UNORM_INT_101010: return "CL_UNORM_INT_101010";
+        case CL_UNORM_INT_101010_2: return "CL_UNORM_INT_101010_2";
         case CL_SIGNED_INT8: return "CL_SIGNED_INT8";
         case CL_SIGNED_INT16: return "CL_SIGNED_INT16";
         case CL_SIGNED_INT32: return "CL_SIGNED_INT32";
@@ -220,6 +221,7 @@ int IsChannelTypeSupported(cl_channel_type type)
         case CL_UNORM_SHORT_565:
         case CL_UNORM_SHORT_555:
         case CL_UNORM_INT_101010:
+        case CL_UNORM_INT_101010_2:
         case CL_SIGNED_INT8:
         case CL_SIGNED_INT16:
         case CL_SIGNED_INT32:
diff --git a/test_common/harness/imageHelpers.cpp b/test_common/harness/imageHelpers.cpp
index e13404b9a5..b63992d75b 100644
--- a/test_common/harness/imageHelpers.cpp
+++ b/test_common/harness/imageHelpers.cpp
@@ -99,7 +99,8 @@ uint32_t get_channel_data_type_size(cl_channel_type channelType)
         case CL_UNORM_SHORT_565:
         case CL_UNORM_SHORT_555: return 2;
 
-        case CL_UNORM_INT_101010: return 4;
+        case CL_UNORM_INT_101010:
+        case CL_UNORM_INT_101010_2: return 4;
 
         case CL_FLOAT: return sizeof(cl_float);
 
@@ -170,6 +171,7 @@ cl_channel_type get_channel_type_from_name(const char *name)
                       { CL_UNORM_SHORT_565, "CL_UNORM_SHORT_565" },
                       { CL_UNORM_SHORT_555, "CL_UNORM_SHORT_555" },
                       { CL_UNORM_INT_101010, "CL_UNORM_INT_101010" },
+                      { CL_UNORM_INT_101010_2, "CL_UNORM_INT_101010_2" },
                       { CL_SIGNED_INT8, "CL_SIGNED_INT8" },
                       { CL_SIGNED_INT16, "CL_SIGNED_INT16" },
                       { CL_SIGNED_INT32, "CL_SIGNED_INT32" },
@@ -934,6 +936,7 @@ float get_max_relative_error(const cl_image_format *format,
         case CL_UNORM_SHORT_565:
         case CL_UNORM_SHORT_555:
         case CL_UNORM_INT_101010:
+        case CL_UNORM_INT_101010_2:
             // Maximum sampling error for round to zero normalization based on
             // multiplication by reciprocal (using reciprocal generated in
             // round to +inf mode, so that 1.0 matches spec)
@@ -1465,6 +1468,15 @@ void read_image_pixel_float(void *imageData, image_descriptor *imageInfo, int x,
             break;
         }
 
+        case CL_UNORM_INT_101010_2: {
+            cl_uint *dPtr = (cl_uint *)ptr;
+            tempData[0] = (float)((dPtr[0] >> 22) & 0x3ff) / (float)1023;
+            tempData[1] = (float)((dPtr[0] >> 12) & 0x3ff) / (float)1023;
+            tempData[2] = (float)(dPtr[0] >> 2 & 0x3ff) / (float)1023;
+            tempData[3] = (float)(dPtr[0] >> 0 & 3) / (float)3;
+            break;
+        }
+
         case CL_FLOAT: {
             float *dPtr = (float *)ptr;
             for (i = 0; i < channelCount; i++) tempData[i] = (float)dPtr[i];
@@ -2730,6 +2742,15 @@ void pack_image_pixel(float *srcVector, const cl_image_format *imageFormat,
                 | (((unsigned int)NORMALIZE(srcVector[2], 1023.f) & 1023) << 0);
             break;
         }
+        case CL_UNORM_INT_101010_2: {
+            cl_uint *ptr = (cl_uint *)outData;
+            ptr[0] =
+                (((unsigned int)NORMALIZE(srcVector[0], 1023.f) & 1023) << 22)
+                | (((unsigned int)NORMALIZE(srcVector[1], 1023.f) & 1023) << 12)
+                | (((unsigned int)NORMALIZE(srcVector[2], 1023.f) & 1023) << 2)
+                | (((unsigned int)NORMALIZE(srcVector[3], 3.f) & 3) << 0);
+            break;
+        }
         case CL_SIGNED_INT8: {
             cl_char *ptr = (cl_char *)outData;
             for (unsigned int i = 0; i < channelCount; i++)
@@ -2892,6 +2913,20 @@ void pack_image_pixel_error(const float *srcVector,
 
             break;
         }
+        case CL_UNORM_INT_101010_2: {
+            const cl_uint *ptr = (const cl_uint *)results;
+
+            errors[0] = ((ptr[0] >> 22) & 1023)
+                - NORMALIZE_UNROUNDED(srcVector[0], 1023.f);
+            errors[1] = ((ptr[0] >> 12) & 1023)
+                - NORMALIZE_UNROUNDED(srcVector[1], 1023.f);
+            errors[2] = ((ptr[0] >> 2) & 1023)
+                - NORMALIZE_UNROUNDED(srcVector[2], 1023.f);
+            errors[3] =
+                ((ptr[0] >> 0) & 3) - NORMALIZE_UNROUNDED(srcVector[3], 3.f);
+
+            break;
+        }
         case CL_SIGNED_INT8: {
             const cl_char *ptr = (const cl_char *)results;
 
diff --git a/test_conformance/images/common.cpp b/test_conformance/images/common.cpp
index 0b2c956cc6..1d7174545b 100644
--- a/test_conformance/images/common.cpp
+++ b/test_conformance/images/common.cpp
@@ -16,12 +16,14 @@
 #include "common.h"
 
 cl_channel_type floatFormats[] = {
-    CL_UNORM_SHORT_565, CL_UNORM_SHORT_555, CL_UNORM_INT_101010,
+    CL_UNORM_SHORT_565,    CL_UNORM_SHORT_555, CL_UNORM_INT_101010,
+    CL_UNORM_INT_101010_2,
 #ifdef CL_SFIXED14_APPLE
     CL_SFIXED14_APPLE,
 #endif
-    CL_UNORM_INT8,      CL_SNORM_INT8,      CL_UNORM_INT16,      CL_SNORM_INT16,
-    CL_FLOAT,           CL_HALF_FLOAT,      (cl_channel_type)-1,
+    CL_UNORM_INT8,         CL_SNORM_INT8,      CL_UNORM_INT16,
+    CL_SNORM_INT16,        CL_FLOAT,           CL_HALF_FLOAT,
+    (cl_channel_type)-1,
 };
 
 cl_channel_type intFormats[] = {
diff --git a/test_conformance/images/kernel_read_write/test_common.cpp b/test_conformance/images/kernel_read_write/test_common.cpp
index a22db19584..328f5ca8e3 100644
--- a/test_conformance/images/kernel_read_write/test_common.cpp
+++ b/test_conformance/images/kernel_read_write/test_common.cpp
@@ -1663,6 +1663,8 @@ int filter_rounding_errors(int forceCorrectlyRoundedWrites,
     if (0 == forceCorrectlyRoundedWrites
         && (imageInfo->format->image_channel_data_type == CL_UNORM_INT8
             || imageInfo->format->image_channel_data_type == CL_UNORM_INT_101010
+            || imageInfo->format->image_channel_data_type
+                == CL_UNORM_INT_101010_2
             || imageInfo->format->image_channel_data_type == CL_UNORM_INT16
             || imageInfo->format->image_channel_data_type == CL_SNORM_INT8
             || imageInfo->format->image_channel_data_type == CL_SNORM_INT16