After watching WWDC 2025 session "Combine Metal 4 machine learning and graphics", I have decided to give it a shot to integrate the latest MTL4MachineLearningCommandEncoder to my existing render pipeline. After a lot of trial and errors, I managed to set up the pipeline and have the app compiled. However, I am now stuck on creating a MTLLibrary with .mtlpackage. Here is the code I have to create a MTLLibrary according the WWDC session https://developer.apple.com/videos/play/wwdc2025/262/?time=550: let coreMLFilePath = bundle.path(forResource: "my_model", ofType: "mtlpackage")! let coreMLURL = URL(string: coreMLFilePath)! do { metalDevice.makeLibrary(URL: coreMLURL) } catch { print("error: \(error)") } With the above code, I am getting error: Error Domain=MTLLibraryErrorDomain Code=1 "Invalid metal package" UserInfo={NSLocalizedDescription=Invalid metal package} What is the correct way to create a MTLLibrary with .mtlpackage? Do I see this error because the .mtlpackage I am using is incorrect? How should I go with debugging this? I'd really appreciate if I could get some help on this as I have been stuck with it for some time now. Thanks in advance!

The code is pretty simple kernel void naive( constant RunParams *param [[ buffer(0) ]], const device float *A [[ buffer(1) ]], // [N, K] device float *output [[ buffer(2) ]], uint2 gid [[ thread_position_in_grid ]]) { uint a_ptr = gid.x * param->K; for (uint i = 0; i < param->K; i++, a_ptr++) { val += A[b_ptr]; } output[ptr] = val; } when uint a_ptr = gid.x * param->K, the code got 150 GFLops when uint a_ptr = gid.y * param->K, the code got 860 GFLops param->K = 256; thread per group: [16, 16] I'd like to understand why the performance is so different, and how can I profile/diagnose this to help with further optimization.

I implemented an EntityAction to change the baseColor tint - and had it working on VisionOS 2.x. import RealityKit import UIKit typealias Float4 = SIMD4<Float> extension UIColor { var float4: Float4 { if cgColor.numberOfComponents == 4, let c = cgColor.components { Float4(Float(c[0]), Float(c[1]), Float(c[2]), Float(c[3])) } else { Float4() } } } struct ColourAction: EntityAction { // MARK: - PUBLIC PROPERTIES let startColour: Float4 let targetColour: Float4 // MARK: - PUBLIC COMPUTED PROPERTIES var animatedValueType: (any AnimatableData.Type)? { Float4.self } // MARK: - INITIATION init(startColour: UIColor, targetColour: UIColor) { self.startColour = startColour.float4 self.targetColour = targetColour.float4 } // MARK: - PUBLIC STATIC FUNCTIONS @MainActor static func registerEntityAction() { ColourAction.subscribe(to: .updated) { event in guard let animationState = event.animationState else { return } let interpolatedColour = event.action.startColour.mixedWith(event.action.targetColour, by: Float(animationState.normalizedTime)) animationState.storeAnimatedValue(interpolatedColour) } } } extension Entity { // MARK: - PUBLIC FUNCTIONS func changeColourTo(_ targetColour: UIColor, duration: Double) { guard let modelComponent = components[ModelComponent.self], let material = modelComponent.materials.first as? PhysicallyBasedMaterial else { return } let colourAction = ColourAction(startColour: material.baseColor.tint, targetColour: targetColour) if let colourAnimation = try? AnimationResource.makeActionAnimation(for: colourAction, duration: duration, bindTarget: .material(0).baseColorTint) { playAnimation(colourAnimation) } } } This doesn't work in VisionOS 26. My current fix is to directly set the material base colour - but this feels like the wrong approach: @MainActor static func registerEntityAction() { ColourAction.subscribe(to: .updated) { event in guard let animationState = event.animationState, let entity = event.targetEntity, let modelComponent = entity.components[ModelComponent.self], var material = modelComponent.materials.first as? PhysicallyBasedMaterial else { return } let interpolatedColour = event.action.startColour.mixedWith(event.action.targetColour, by: Float(animationState.normalizedTime)) material.baseColor.tint = UIColor(interpolatedColour) entity.components[ModelComponent.self]?.materials[0] = material animationState.storeAnimatedValue(interpolatedColour) } } So before I raise this as a bug, was I doing anything wrong in the former version and got lucky? Is there a better approach?

Hi everyone, I’ve been developing a custom, end-to-end 3D rendering engine called Crescent from scratch using C++20 and Metal-cpp (targeting macOS and visionOS). My primary goal is to build a zero-bottleneck, GPU-driven pipeline that maximizes the potential of Apple Silicon’s Unified Memory and TBDR architecture. While the fundamental systems are stable, I am looking for architectural feedback from Metal framework engineers regarding specific synchronization and latency challenges. Current Core Implementations: GPU-Driven Instance Culling: High-performance occlusion culling using a Hierarchical Z-Buffer (HZB) approach via Compute Shaders. Clustered Forward Shading: Support for high-count dynamic lights through view-space clustering. Temporal Stability: Custom TAA with history rejection and Motion Blur resolve. Asset Infrastructure: Robust GUID-based scene serialization and a JSON-driven ECS hierarchy. The Architectural Challenge: I am currently seeing slight synchronization overhead when generating the HZB mip-chain. On Apple Silicon, I am evaluating the cost of encoder transitions versus cache-friendly barriers. && m_hzbInitPipeline && m_hzbDownsamplePipeline && !m_hzbMipViews.empty(); if (canBuildHzb) { MTL::ComputeCommandEncoder* hzbInit = commandBuffer->computeCommandEncoder(); hzbInit->setComputePipelineState(m_hzbInitPipeline); hzbInit->setTexture(m_depthTexture, 0); hzbInit->setTexture(m_hzbMipViews[0], 1); if (m_pointClampSampler) { hzbInit->setSamplerState(m_pointClampSampler, 0); } else if (m_linearClampSampler) { hzbInit->setSamplerState(m_linearClampSampler, 0); } const uint32_t hzbWidth = m_hzbMipViews[0]->width(); const uint32_t hzbHeight = m_hzbMipViews[0]->height(); const uint32_t threads = 8; MTL::Size tgSize = MTL::Size(threads, threads, 1); MTL::Size gridSize = MTL::Size((hzbWidth + threads - 1) / threads * threads, (hzbHeight + threads - 1) / threads * threads, 1); hzbInit->dispatchThreads(gridSize, tgSize); hzbInit->endEncoding(); for (size_t mip = 1; mip < m_hzbMipViews.size(); ++mip) { MTL::Texture* src = m_hzbMipViews[mip - 1]; MTL::Texture* dst = m_hzbMipViews[mip]; if (!src || !dst) { continue; } MTL::ComputeCommandEncoder* downEncoder = commandBuffer->computeCommandEncoder(); downEncoder->setComputePipelineState(m_hzbDownsamplePipeline); downEncoder->setTexture(src, 0); downEncoder->setTexture(dst, 1); const uint32_t mipWidth = dst->width(); const uint32_t mipHeight = dst->height(); MTL::Size downGrid = MTL::Size((mipWidth + threads - 1) / threads * threads, (mipHeight + threads - 1) / threads * threads, 1); downEncoder->dispatchThreads(downGrid, tgSize); downEncoder->endEncoding(); } if (m_instanceCullHzbPipeline) { dispatchInstanceCulling(m_instanceCullHzbPipeline, true); } } My Questions: Encoder Synchronization: Would you recommend moving this loop into a single ComputeCommandEncoder using MTLBarrier between dispatches to maintain L2 cache residency, or is the overhead of separate encoders negligible for depth-downsampling on TBDR? visionOS Bindless Latency: For stereo rendering on visionOS, what are the best practices for managing MTL4ArgumentTable updates at 90Hz+? I want to ensure that updating bindless resources for each eye doesn't introduce unnecessary CPU-to-GPU latency. Memory Management: Are there specific hints for Memoryless textures that could be applied to intermediate HZB levels to save bandwidth during this process? I’ve attached a screenshot of a scene rendered with the engine (PBR, SSR, and IBL).

We are developing a standalone AI avatar application for hospital reception kiosks using Mac mini (M2/M4). The app runs on SwiftUI + RealityKit, displays on a 75-inch monitor, and utilizes a USB-connected 4K camera and external sensors (LiDAR/mmWave). We have several technical concerns regarding the transition from iPadOS to macOS. Could you please provide insights on the following? ARKit/Vision Framework on macOS with External Camera On iPadOS, ARKit provides robust Face Tracking. On macOS with an external USB 4K camera: Can we achieve real-time face tracking (expression/gaze/depth) with Vision framework or ARKit comparable to iPadOS performance? Are there any specific limitations for accessing the Neural Engine via Vision framework for real-time 4K video analysis on macOS? Accessing External Hardware (LiDAR/Sensors) in Sandbox We plan to connect external LiDAR and mmWave sensors (e.g., Akara) via USB/Bluetooth. Is it feasible to communicate with these custom drivers/devices within the App Sandbox environment? Would DriverKit be required, or can we use standard serial communication APIs? On-Device LLM (MLX) & Thermals We intend to run a local LLM (e.g., Llama 3 using MLX framework) for offline conversation, alongside 3D rendering. With the M2/M4 Mac mini fan design, is there a risk of thermal throttling during 10+ hours of continuous operation (simultaneous CoreML + 3D rendering)? Is the Mac Studio recommended over the Mac mini for this thermal profile? Long-running Speech API Are there any known issues (memory leaks, API limits) when using Spherch framework and AVSpeechSynthesizer continuously for over 10 hours daily? 3D Display Output Are there any macOS constraints for rendering a SwiftUI window in a specific 3D format (e.g., Side-by-Side) and outputting it via HDMI to a 3D digital signage display (fixed refresh rate/resolution)? Thank you for your assistance.

I have an odd bug, if I use initWithFrame as the init routine for NSView subclass that uses layers I don't see this bug. But if I embedded this view into a storyboard with a .nib file and use initWithCoder, I need to return true on (BOOL) contentsAreFlipped From the NSView subclass If I don't the CALayer actually renders from 0,0 from the view upwards and off the window. The frame sizes for the NSView and the CALayer are good.. when I see them in updateLayer. Obviously I have a fix.. but I would like to understand why.

I use unity 2020.3.48f1 to develop a game; trying to implement Apple Services integration I use Apple unity plugins(https://github.com/apple/unityplugins) Using latest version of unity plugins I getting error in Unity project after plugin import It say "Not allowed platform VisionOS" When I tryed to use older version of the plugins I getting error on runtime when calling "var fetchItemsResponse = await GKLocalPlayer.Local.FetchItems();" in line 42 it drop EXC_BAD_ACCESS(code=257, address=0x0000...) error I tryed to use different commits from official repositorys and even custom branches of apple unity plugins like (https://github.com/muZZkat/unityplugins/tree/muzzkat/fix-fetch-items) but it did not help There is whole my script which trying to use apple unuity plugins using System.Threading.Tasks; using UnityEngine; using System.Collections; using System; using Apple.GameKit; using UnityEngine.UI; public class TheScript : MonoBehaviour { [SerializeField] InputField otp; string Signature; string TeamPlayerID; string Salt; string PublicKeyUrl; string Timestamp; void Start() { StartCoroutine(Call()); } private IEnumerator Call() { yield return new WaitForSeconds(5); Login(); } public async Task Login() { otp.text += $"Loginig... "; if (!Apple.GameKit.GKLocalPlayer.Local.IsAuthenticated) { try { var player = await GKLocalPlayer.Authenticate(); var localPlayer = GKLocalPlayer.Local; TeamPlayerID = localPlayer.TeamPlayerId; var fetchItemsResponse = await GKLocalPlayer.Local.FetchItems(); Signature = Convert.ToBase64String(fetchItemsResponse.GetSignature()); PublicKeyUrl = fetchItemsResponse.PublicKeyUrl; otp.text += $"Team Player ID: {TeamPlayerID} "; otp.text += $"PublicKeyUrl: {PublicKeyUrl} "; } catch(Exception e) { otp.text += $"Error: " + e.Message; } } else { Debug.Log("AppleGameCenter player already logged in."); } } async Task SignInWithAppleGameCenterAsync(string signature, string teamPlayerId, string publicKeyURL, string salt, ulong timestamp) { } }

I have a game built in Unreal Engine 5.6 which uses tilt motion controls to rotate an object. I've restricted the app to only run in portrait for iPhone, and everything works fine, however for iPad I've had a few issues relating to multitasking and I can't seem to solve it. Forcing the app to portrait only still allows the app to run in landscape mode, but shows black bars either side of the game, and the axes for the motion controls are incorrect. X becomes Y and Y becomes X, and there's no way for my app to know which orientation it is because the container is still technically portrait. Allowing my game to run in all orientations makes the whole app more presentable, it doesn't add black bars and the game is still functional and I'm able to map the controls correctly because the game knows it's landscape rather than portrait. The problem with allowing my app to run in landscape mode is if multitasking is enabled on the ipad, you can resize the app to be portrait, and then I run into the same problem again where the game thinks it's portrait mode and all of the axes are wrong again. I tried getting the true orientation of the device rather than the scene, but the game is intended to be played flat so instead of returning the orientation of the OS the orientation is FaceUp, which doesn't help. I need to either disable multitasking or find a way of getting the orientation of the OS (not the scene or the device). I haven't found how to get the OS orientation so I've been trying to disable multitasking. I've got Requires Fullscreen true and UIApplicationSupportsMultipleScreens false in my info.plist but my iPad still seems to allow the window to be resized in landscape view. Opening the IOS workspace of my project Requires Fullscreen is ticked but under that it says "Supports Multiple Windows" and the arrow button next to it takes my to my info.plist values, but no indication of how I can change it. I'm using Unreal Engine 5.6 and Xcode 16.0. Xcode is old I know, but this version of unreal engine doesn't seem to support any newer.

Hi, I'm trying to add game center challenges and activities to an already live game, but they are not appearing in game for testing, GameCenter, or the Games app. I know the game is setup with GameKit entitlements since this is a live game and it has working leaderboards and achievements. I've updated to Tahoe beta 8, added a challenge and activity on app store connect, added that to a new distribution and added that distribution to 'Add for Review' I'm using Unity and the Apple Unity plugin Not sure what other steps I'm missing Thanks

I do not understand how offline leaderboard submission is supposed to work in Game Kit: While the documentation briefly states that offline submission is supported, how is that even possible when you first have to fetch a leaderboard object in order to then call its submitScore function? How can I get the leaderboard object in the first place when offline? Can anyone enlighten me how this works? Or maybe point me to some relevant documentation?

Hello! I have a question about how thread groups work with tile shading. When running "traditional" compute, I get to choose both thread group size and the grid size. However, when using tile shading kernel I only have dispatchThreadsPerTile method - this controls how many threads will be ran in each tile. So far so good, but what about thread groups? The examples in video "Tile Shading on A11" seem to suggest that there will be only one thread group per tile. In the video, [[thread_index_in_threadgroup]] is called "local_id" and it is used to access the image block. I assume this is the default configuration. So when one does the following: Creates MTLRenderPassDescriptor with tileWidth set to W and tileHeight set to H Fires up the tile shading kernel using dispatchThreadsPerTile with MTLSize size = { W, H, 1 } I understand that the result is 1-to-1 mapping between the tile "pixels" and kernel threads. Now, what I would like to do is to have more than one thread group there. I want this for performance reasons: I have a certain compute kernel which I know executes very well with small thread group size. In fact, { 32, 1, 1 } seems to be the fastest. My understanding is that even if I set tile size to 16x16, and so I am executing 256 threads there, there will only be one SIMD group active in a thread group. Meaning that this SIMD group has to execute 8 times over the tile. Is it possible somehow? Or perhaps the limitations of the API are pointing at the limitations of hardware itself, and if I want to execute with SIMD group sized thread groups I have to use "traditional" compute encoder? Will be grateful for help. Michał

after launching a nearby exoerience on quick look or inside our app, all the user in the group watch sometimes teh model blinking abd becoming transparet... ... just one user hasnt the issue, either the one who launched shareplay or the user who force align the immersive space in front weird

I have a question I guess more for the Apple team. But why are there no totally 3D experiences for the Vision Pro lineup? I know they have given us tools to implement unity 3D games into iPhone and I guess you can also build it in RealityKit. But why at this moment are 3D games limited to just iPad and iPhone and can't you bring that into Vision Pro? Just to explain. When I say a totally 3D game, I mean games like Gorn. I mean the Vision Pro is definitely powerful enough, but it just feels limited to tabletop games and AR games. Is this something Apple is thinking about implementing?

Hello I would like to know how to combine 2 animations with RealityKit (one animation for the arms and one for the legs for example) I saw this apple demo that seems to explain it but I don't understand at all how to do it... Thanks

Hello Everyone I am new here, I am testing game center integration and using a development build of my IOS game. I have set up a couple of achievements in app store connect, but when I trigger them in the game then they do not unlock or show up. Okay so i am signed into the game center with a sandbox account on a test advice. Is there anything else I need to configure, or do achievements usually only work after the game is released?. I will appreciate any guidance… Thanks in Advance!!!

Hi, I would like clarification on whether the new hover effects feature introduced in vision os 26 supported pinch gestures through the psvr 2 controllers. In your sample application, I was not able to confirm that this was working. Only pinch clicking with my hands worked. Pulling the trigger on the controller whilst looking at a 3d object did not activate the hover effect spatial event in the sample application. (The object is showing the highlight though) This is inconsistent with hover effect behavior with psvr2 controllers on swift ui views, where the trigger press does count as a button click. The sample I used was this one: https://developer.apple.com/documentation/compositorservices/rendering_hover_effects_in_metal_immersive_apps

Problem Summary After upgrading to iOS 26.1 and 26.2, I'm experiencing a particle positioning bug in RealityKit where ParticleEmitterComponent particles render at an incorrect offset relative to their parent entity. This behavior does not occur on iOS 18.6.2 or earlier versions, suggesting a regression introduced in the newer OS builds. Environment Details Operating System: iOS 26.1 & iOS 26.2 Framework: RealityKit Xcode Version: 16.2 (16C5032a) Expected vs. Actual Behavior Expected: Particles should render at the position of the entity to which the ParticleEmitterComponent is attached, matching the behavior on iOS 18.6.2 and earlier. Actual: Particles appear away from their parent entity, creating a visual misalignment that breaks the intended AR experience. Steps to Reproduce Create or open an AR application with RealityKit that uses particle components Attach a ParticleEmitterComponent to an entity via a custom system Run the application on iOS 26.1 or iOS 26.2 Observe that particles render at an offset position away from the entity Minimal Code Example Here's the setup from my test case: Custom Component & System: struct SparkleComponent4: Component {} class SparkleSystem4: System { static let query = EntityQuery(where: .has(SparkleComponent4.self)) required init(scene: Scene) {} func update(context: SceneUpdateContext) { for entity in context.scene.performQuery(Self.query) { // Only add once if entity.components.has(ParticleEmitterComponent.self) { continue } var newEmitter = ParticleEmitterComponent() newEmitter.mainEmitter.color = .constant(.single(.red)) entity.components.set(newEmitter) } } } AR Setup: let material = SimpleMaterial(color: .gray, roughness: 0.15, isMetallic: true) let model = Entity() model.components.set(ModelComponent(mesh: boxMesh, materials: [material])) model.components.set(SparkleComponent4()) model.position = [0, 0.05, 0] model.name = "MyCube" let anchor = AnchorEntity(.plane(.horizontal, classification: .any, minimumBounds: [0.2, 0.2])) anchor.addChild(model) arView.scene.addAnchor(anchor) Questions for the Community Has anyone else encountered this particle positioning issue after updating to iOS 26.1/26.2? Are there known workarounds or configuration changes to ParticleEmitterComponent that restore correct positioning? Is this a confirmed bug, or could there be a change in coordinate system handling or transform inheritance that I'm missing? Additional Information I've already submitted this issue via Feedback Assistant(FB21346746)

App Storeにある『浮遊時計 Premium』は1Hzごとか10Hzごと、または3段階以上のリフレッシュレート計測はできますか？

I'm trying to use MTLBinaryArchive. I collected a BinaryArchive from one device and used metal-tt to translate it for all supported iPhone devices, ranging from iPhone 7 Plus to iPhone 16. However, this BinaryArchive is quite large, around 1.5GB uncompressed, and about 500MB compressed in the IPA. I'm wondering how to address the size issue. I watched the WWDC 2022 video, which mentioned that the operating system or app installation process would handle compatibility. Does this compatibility support different GPU chips? I tried installing an IPA with a BinaryArchive collected only from an iPhone 12 on an iPhone 13, but the BinaryArchive didn't take effect. I also saw that Apple supports App Thinning. However, it seems that resources in the Asset Catalog cannot be accessed via URL, and creating an MTLBinaryArchive requires a URL. Is it possible for MTLBinaryArchive to be distributed through App Thinning? The WWDC 2022 video also mentioned using the -Os optimization flag to reduce size. Can this give an estimate of how much compression it would achieve? Are there any methods to solve the BinaryArchive size issue without impacting performance?

I'm using RealityView in my iOS game mxied with SwiftUI. For the following 2 example usages, the simulator will only render the first RealityView, and the second one is either super laggy or show a black model. Running on the real device is all good, just simualtor has this issue. Have a TabView and each tab has a RealityView. Have a root view and detail view connected via a push navigation, both root and detail have a RealityView. In the Simulator, the second RealityView is going to be very choppy and basically unusable, but on a real iPhone everything looks great. Is this a known simulator issue or I did something bad?