Fix provider device selection

Author: saddam213

Directory.Build.props

@@ -1,6 +1,6 @@
 <Project>
   <PropertyGroup>
-    <Version>0.2.0</Version>
+    <Version>0.2.1</Version>
     <Company>TensorStack</Company>
     <Copyright>TensorStack - 2025</Copyright>
     <RepositoryUrl>https://github.com/TensorStack-AI/TensorStack</RepositoryUrl>

TensorStack.Providers.RyzenAI/Provider.cs

@@ -1,4 +1,4 @@
-// Copyright (c) TensorStack. All rights reserved.
+// Copyright (c) TensorStack, Advanced Micro Devices. All rights reserved.
 // Licensed under the Apache 2.0 License.
 using Microsoft.ML.OnnxRuntime;
 using System.Collections.Generic;
@@ -8,11 +8,13 @@
 
 namespace TensorStack.Providers
 {
+    /// <summary>
+    /// RyzenAI NPU provider with DirectML GPU fallback
+    /// </summary>
     public static class Provider
     {
         private static bool _isInitialized;
-        private const string _providerName = "DMLExecutionProvider";
-
+        private const string _providerName = "RyzenAIExecutionProvider";
 
         /// <summary>
         /// Initializes the Provider 
@@ -23,7 +25,7 @@ public static void Initialize()
                 return;
 
             _isInitialized = true;
-            DeviceManager.Initialize(_providerName);
+            DeviceManager.Initialize("DMLExecutionProvider");
         }
 
 
@@ -37,7 +39,7 @@ public static void Initialize(EnvironmentCreationOptions environmentOptions)
                 return;
 
             _isInitialized = true;
-            DeviceManager.Initialize(environmentOptions, _providerName);
+            DeviceManager.Initialize(environmentOptions, "DMLExecutionProvider");
         }
 
 
@@ -73,6 +75,9 @@ public static Device GetDevice()
         /// <param name="deviceType">Type of the device.</param>
         public static Device GetDevice(DeviceType deviceType)
         {
+            if (deviceType == DeviceType.NPU)
+                return GetDevices().FirstOrDefault(x => x.Type == DeviceType.GPU);
+
             return GetDevices().FirstOrDefault(x => x.Type == deviceType);
         }
 
@@ -84,12 +89,15 @@ public static Device GetDevice(DeviceType deviceType)
         /// <param name="deviceId">The device identifier.</param>
         public static Device GetDevice(DeviceType deviceType, int deviceId)
         {
+            if (deviceType == DeviceType.NPU)
+                return GetDevices().FirstOrDefault(x => x.Type == DeviceType.GPU && x.DeviceId == deviceId);
+
             return GetDevices().FirstOrDefault(x => x.Type == deviceType && x.DeviceId == deviceId);
         }
 
 
         /// <summary>
-        /// Gets the DirectML provider this DeviceType.
+        /// Gets the RyzenAI provider this DeviceType.
         /// </summary>
         /// <param name="optimizationLevel">The optimization level.</param>
         public static ExecutionProvider GetProvider(GraphOptimizationLevel optimizationLevel = GraphOptimizationLevel.ORT_ENABLE_ALL)
@@ -99,7 +107,7 @@ public static ExecutionProvider GetProvider(GraphOptimizationLevel optimizationL
 
 
         /// <summary>
-        /// Gets the DirectML provider this DeviceType.
+        /// Gets the RyzenAI provider this DeviceType.
         /// </summary>
         /// <param name="deviceType">Type of the device.</param>
         /// <param name="optimizationLevel">The optimization level.</param>
@@ -110,7 +118,7 @@ public static ExecutionProvider GetProvider(DeviceType deviceType, GraphOptimiza
 
 
         /// <summary>
-        /// Gets the DirectML provider this DeviceType, DeviceId.
+        /// Gets the RyzenAI provider this DeviceType, DeviceId.
         /// </summary>
         /// <param name="deviceType">Type of the device.</param>
         /// <param name="deviceId">The device identifier.</param>
@@ -122,25 +130,23 @@ public static ExecutionProvider GetProvider(DeviceType deviceType, int deviceId,
 
 
         /// <summary>
-        /// Gets the DirectML provider for this Device.
+        /// Gets the RyzenAI provider for this Device.
         /// </summary>
         /// <param name="device">The device.</param>
         /// <param name="optimizationLevel">The optimization level.</param>
         public static ExecutionProvider GetProvider(this Device device, GraphOptimizationLevel optimizationLevel = GraphOptimizationLevel.ORT_ENABLE_ALL)
         {
             if (device == null)
                 return default;
-            else if (device.Type == DeviceType.NPU)
-                return CreateRyzenProvider(device.DeviceId, optimizationLevel);
             else if (device.Type == DeviceType.CPU)
                 return CreateCPUProvider(optimizationLevel);
 
-            return CreateDMLProvider(device.DeviceId, optimizationLevel);
+            return CreateRyzenProvider(device.DeviceId, optimizationLevel);
         }
 
 
         /// <summary>
-        /// Gets the DirectML provider for this DeviceId.
+        /// Gets the RyzenAI provider for this DeviceId.
         /// </summary>
         /// <param name="deviceId">The device identifier.</param>
         /// <param name="optimizationLevel">The optimization level.</param>
@@ -166,29 +172,6 @@ private static ExecutionProvider CreateRyzenProvider(int deviceId, GraphOptimiza
         }
 
 
-        /// <summary>
-        /// Gets the DirectML provider for this DeviceId.
-        /// </summary>
-        /// <param name="deviceId">The device identifier.</param>
-        /// <param name="optimizationLevel">The optimization level.</param>
-        private static ExecutionProvider CreateDMLProvider(int deviceId, GraphOptimizationLevel optimizationLevel = GraphOptimizationLevel.ORT_ENABLE_ALL)
-        {
-            var memoryInfo = new OrtMemoryInfo(OrtMemoryInfo.allocatorCPU, OrtAllocatorType.DeviceAllocator, deviceId, OrtMemType.Default);
-            return new ExecutionProvider(_providerName, memoryInfo, configuration =>
-            {
-                var sessionOptions = new SessionOptions
-                {
-                    GraphOptimizationLevel = optimizationLevel
-                };
-
-                sessionOptions.AddSessionConfigEntries(configuration.SessionOptions);
-                sessionOptions.AppendExecutionProvider_DML(deviceId);
-                sessionOptions.AppendExecutionProvider_CPU();
-                return sessionOptions;
-            });
-        }
-
-
         /// <summary>
         /// Gets the CPU provider.
         /// </summary>

TensorStack.Providers.RyzenAI/TensorStack.Providers.RyzenAI.csproj

@@ -3,7 +3,8 @@
   <PropertyGroup>
     <TargetFrameworks>net9.0-windows10.0.17763.0;net10.0-windows10.0.17763.0</TargetFrameworks>
     <PlatformTarget>x64</PlatformTarget>
-    <Copyright>Advanced Micro Devices, TensorStack - 2025</Copyright>
+    <Authors>Advanced Micro Devices, TensorStack</Authors>
+    <Copyright>Advanced Micro Devices - 2025, TensorStack - 2025</Copyright>
 
     <!--DirectML.dll content warning-->
     <NoWarn>$(NoWarn);NU5100</NoWarn>