diff --git a/compiler/src/iree/compiler/Codegen/LLVMCPU/KernelDispatch.cpp b/compiler/src/iree/compiler/Codegen/LLVMCPU/KernelDispatch.cpp
index 9ce86e9a553d..c485b7f99f18 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMCPU/KernelDispatch.cpp
+++ b/compiler/src/iree/compiler/Codegen/LLVMCPU/KernelDispatch.cpp
@@ -30,6 +30,7 @@
 #include "mlir/Dialect/Utils/IndexingUtils.h"
 #include "mlir/Dialect/Utils/StaticValueUtils.h"
 #include "mlir/IR/Matchers.h"
+#include "mlir/IR/TypeUtilities.h"
 #include "mlir/Interfaces/FunctionInterfaces.h"
 #include "mlir/Transforms/GreedyPatternRewriteDriver.h"
 
@@ -1179,7 +1180,11 @@ getDefaultMatmulVectorSizes(linalg::LinalgOp op, int64_t vectorSize,
                             SmallVectorImpl<bool> &scalableSizeFlags) {
   auto targetAttr = IREE::HAL::ExecutableTargetAttr::lookup(op);
   if (isX86(targetAttr)) {
-    sizes.append({8, 32, 16});
+    if (hasAVX512fFeature(targetAttr)) {
+      sizes.append({8, 32, 16});
+    } else {
+      sizes.append({1, 1, vectorSize});
+    }
     return;
   }
 
diff --git a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/pipeline_tests.mlir b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/pipeline_tests.mlir
index 06883c2afbd6..3f3ea8499cd5 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/pipeline_tests.mlir
+++ b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/pipeline_tests.mlir
@@ -46,7 +46,7 @@ module {
 //       CHECK:    memref.store
 
 // -----
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
 #map = affine_map<(d0, d1) -> (d0, d1)>
 module {
   func.func @peel_partially_unaligned_matmul() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
@@ -92,7 +92,7 @@ module {
 //       CHECK:     arith.maximumf {{.*}} : vector<
 
 // -----
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
 module {
   func.func @batch_matmul_dynamic() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -193,7 +193,7 @@ module {
 //       CHECK:    arith.cmpf olt, %{{.+}}, %{{.+}} : vector<4x4xf32>
 
 // -----
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu = "generic", cpu_features = "", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu = "generic", cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 64 : index, target_triple = "x86_64-none-elf"}>
 #map = affine_map<(d0, d1) -> (d0, d1)>
 #map1 = affine_map<(d0, d1) -> (d1)>
 module {
diff --git a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_lowering_strategy_without_distribution.mlir b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_lowering_strategy_without_distribution.mlir
index 834647e3dd9e..f978d3332f50 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_lowering_strategy_without_distribution.mlir
+++ b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_lowering_strategy_without_distribution.mlir
@@ -18,8 +18,8 @@ module {
 }
 
 
-//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[0, 0], [8, 32], [0, 0], [0, 0]]>
-//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[0, 0, 0], [0, 0, 0], [0, 0, 0], [8, 32, 0], [0, 0, 16], [0, 0, 0]]>
+//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[0, 0], [1, 1], [0, 0], [0, 0]]>
+//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[0, 0, 0], [0, 0, 0], [0, 0, 0], [1, 1, 0], [0, 0, 4], [0, 0, 0]]>
 //  CHECK-DAG: #[[TRANSLATION:.+]] = #iree_codegen.translation_info<CPUDoubleTilingExpert, {{\{}}enable_loop_peeling}>
 //      CHECK: func.func @matmul_static()
 // CHECK-SAME:     translation_info = #[[TRANSLATION]]
diff --git a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_x86_64_lowering_strategy.mlir b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_x86_64_lowering_strategy.mlir
index 43b3ddb9971d..53e33f973db7 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_x86_64_lowering_strategy.mlir
+++ b/compiler/src/iree/compiler/Codegen/LLVMCPU/test/select_x86_64_lowering_strategy.mlir
@@ -1,6 +1,6 @@
 // RUN: iree-opt --pass-pipeline='builtin.module(iree-llvmcpu-select-lowering-strategy)' --split-input-file %s | FileCheck %s
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
 module {
   func.func @matvec_static() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -27,7 +27,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
 module {
   func.func @matvec_dynamic() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %c0 = arith.constant 0 : index
@@ -62,7 +62,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
 module {
   func.func @dot_static() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -89,7 +89,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
 module {
   func.func @dot_dynamic() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %c0 = arith.constant 0 : index
@@ -244,7 +244,7 @@ module {
 
 // -----
 
-#executable_target_system_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "system-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
+#executable_target_system_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "system-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-unknown-linux-gnu"}>
 module {
   func.func @matmul_partially_peel() attributes {hal.executable.target = #executable_target_system_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -586,7 +586,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
 module {
   func.func @matmul_static() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -665,7 +665,7 @@ module {
   }
 }
 
-//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[64, 64, 0], [64, 64, 0], [0, 0, 0], [8, 32, 0], [0, 0, 16], [0, 0, 0]]>
+//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[64, 64, 0], [64, 64, 0], [0, 0, 0], [1, 1, 0], [0, 0, 4], [0, 0, 0]]>
 //  CHECK-DAG: #[[TRANSLATION:.+]] = #iree_codegen.translation_info<CPUDoubleTilingExpert, {{\{}}enable_loop_peeling}>
 //      CHECK: func.func @matmul_i8_i8_i32_static()
 // CHECK-SAME:     translation_info = #[[TRANSLATION]]
@@ -674,7 +674,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
 module {
   func.func @gemm_unit_N() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %c0 = arith.constant 0 : index
@@ -701,7 +701,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
 module {
   func.func @gemm_unit_M_unit_N() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %c0 = arith.constant 0 : index
@@ -727,7 +727,7 @@ module {
 
 // -----
 
-#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
+#executable_target_embedded_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "embedded-elf-x86_64", {cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", native_vector_size = 16 : index, target_triple = "x86_64-none-elf"}>
 module {
   func.func @matmul_odd() attributes {hal.executable.target = #executable_target_embedded_elf_x86_64_} {
     %cst = arith.constant 0.000000e+00 : f32
@@ -1351,7 +1351,7 @@ module {
   }
 }
 
-//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[64, 48, 0], [64, 48, 0], [0, 0, 0], [8, 32, 0], [0, 0, 16], [0, 0, 0]]>
+//  CHECK-DAG: #[[CONFIG:.+]] = #iree_codegen.lowering_config<tile_sizes = {{\[}}[64, 16, 0], [64, 16, 0], [0, 0, 0], [1, 1, 0], [0, 0, 4], [0, 0, 0]]>
 //  CHECK-DAG: #[[TRANSLATION:.+]] = #iree_codegen.translation_info<CPUDoubleTilingExpert, {{\{}}enable_loop_peeling}>
 //      CHECK: func.func @quant_model()
 // CHECK-SAME:     translation_info = #[[TRANSLATION]]
@@ -1384,7 +1384,7 @@ module {
 
 // -----
 
-#executable_target_system_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "system-elf-x86_64", {cpu = "cascadelake", cpu_features = "", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", link_embedded = false, native_vector_size = 64 : index, target_triple = "x86_64-unknown-linux-gnu", ukernels = false}>
+#executable_target_system_elf_x86_64_ = #hal.executable.target<"llvm-cpu", "system-elf-x86_64", {cpu = "cascadelake", cpu_features = "+avx512f", data_layout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", link_embedded = false, native_vector_size = 64 : index, target_triple = "x86_64-unknown-linux-gnu", ukernels = false}>
 #map = affine_map<(d0, d1) -> (d0, d1)>
 #map1 = affine_map<(d0, d1) -> (d0)>
 module {