I'm running into a persistent visual issue while deploying a floral corridor scene to Apple Vision Pro using Unity 6.0 with URP and Metal. The issue only appears on the Vision Pro device — everything looks fine in the Unity Editor. Issue Description When the frame rate drops to around 60–70 FPS, noticeable distortion artifacts appear around the edges of foliage models. It seems like the background meshes (behind the plants) get warped and leak through the edges of the foliage. Although this is most visible around the leaves, even solid objects like standard URP wall or box models show distorted edges when the issue occurs. All the foliage uses Opaque or Alpha Clipping materials. Things I've Tried Changing the foliage materials to Transparent mode —distortion around edges disappears, but using Transparent for a large number of foliage assets is not ideal for performance or sorting complexity. Reducing the number of foliage objects — with only a few plants in the scene and the frame rate staying around 100 FPS, the distortion disappears. However, this isn’t a practical solution for a full environment. Possible Cause? I came across this note in the Unity documentation: "Ensure depth-buffer for each pixel is non-zero - on visionOS, the depth buffer is used for reprojection. To ensure visual effects like skyboxes and shaders are displayed beautifully, ensure that some value is written to the depth for each pixel." Could this be related to the issue? Is it possible that Alpha Clipping with low pixel coverage leads to some pixels not writing to the depth buffer, which then causes problems during Vision Pro’s reprojection or foveated rendering? However, even when I disable Alpha Clipping entirely, the distortion issue still persists, so it may not be solely caused by clipping itself. Project Setup Unity 6.0 (URP) Depth Texture: Enable Using Metal as the graphics backend Running on real Vision Pro hardware (not simulator) Any advice on how to avoid these distortion issues on Vision Pro would be greatly appreciated. Thanks!

hi everyone, 我们发现了一个和Metal相关崩溃。应用中使用了Metal相关的接口，在进行性能测试时，打开了设置-开发者-显示HUD图形。运行应用后，正常展示HUD，但应用很快发生了崩溃，日志主要信息如下： Incident Identifier: 1F093635-2DB8-4B29-9DA5-488A6609277B CrashReporter Key: 233e54398e2a0266d95265cfb96c5a89eb3403fd Hardware Model: iPhone14,3 Process: waimai [16584] Path: /private/var/containers/Bundle/Application/CCCFC0AE-EFB8-4BD8-B674-ED089B776221/waimai.app/waimai Identifier: Version: 61488 (8.53.0) Code Type: ARM-64 Parent Process: ? [1] Date/Time: 2025-06-12 14:41:45.296 +0800 OS Version: iOS 18.0 (22A3354) Report Version: 104 Monitor Type: Mach Exception Exception Type: EXC_BAD_ACCESS (SIGBUS) Exception Codes: KERN_PROTECTION_FAILURE at 0x000000014fffae00 Crashed Thread: 57 Thread 57 Crashed: 0 libMTLHud.dylib esfm_GenerateTriangesForString + 408 1 libMTLHud.dylib esfm_GenerateTriangesForString + 92 2 libMTLHud.dylib Renderer::DrawText(char const*, int, unsigned int) + 204 3 libMTLHud.dylib Overlay::onPresent(id<CAMetalDrawable>) + 1656 4 libMTLHud.dylib CAMetalDrawable_present(void (*)(), objc_object*, objc_selector*) + 72 5 libMTLHud.dylib invocation function for block in void replaceMethod<void>(objc_class*, objc_selector*, void (*)(void (*)(), objc_object*, objc_selector*)) + 56 6 Metal __45-[_MTLCommandBuffer presentDrawable:options:]_block_invoke + 104 7 Metal MTLDispatchListApply + 52 8 Metal -[_MTLCommandBuffer didScheduleWithStartTime:endTime:error:] + 312 9 IOGPU IOGPUNotificationQueueDispatchAvailableCompletionNotifications + 136 10 IOGPU __IOGPUNotificationQueueSetDispatchQueue_block_invoke + 64 11 libdispatch.dylib _dispatch_client_callout4 + 20 12 libdispatch.dylib _dispatch_mach_msg_invoke + 464 13 libdispatch.dylib _dispatch_lane_serial_drain + 368 14 libdispatch.dylib _dispatch_mach_invoke + 456 15 libdispatch.dylib _dispatch_lane_serial_drain + 368 16 libdispatch.dylib _dispatch_lane_invoke + 432 17 libdispatch.dylib _dispatch_lane_serial_drain + 368 18 libdispatch.dylib _dispatch_lane_invoke + 380 19 libdispatch.dylib _dispatch_root_queue_drain_deferred_wlh + 288 20 libdispatch.dylib _dispatch_workloop_worker_thread + 540 21 libsystem_pthread.dylib _pthread_wqthread + 288 我们测试了几个不同的机型，只有iPhone 13 Pro Max会发生崩溃。 Q1：为什么会发生这个崩溃？ Q2：相同的逻辑，为什么仅在iPhone 13 Pro Max机型上出现崩溃？ 期待您的解答。

Hi there, I'm wondering if it's possible under iOS 28 developer beta to enable MetalFX scaling info with '{"MTL_HUD_ENABLED": "1" for my App. This information has been added to Mac, but looks to be absent on iPhone / iPad

Hello, I'm tracking down a bug where useResource doesn't seem to apply proper synchronization when a resource is produced by the render pass then consumed by the compute pass, but when I use MTLFence between the to signal and wait between the render/compute encoders, the artifact goes away. The resource is created with MTLHazardTrackingModeTracked and useResource is called on the compute encoder after the render pass. Metal API Validation doesn't report any warnings/errors. Am I misunderstanding the difference between the two APIs? I dug through the Metal documentation and it looks like useResource should handle synchronization given the resource has MTLHazardTrackingModeTracked but on the other hand, MTLFence should be used to ensure proper synchronization between command encoders. Can someone can clarify the difference between the two APIs and when to use them.

I have really enjoyed looking through the code and videos related to Metal 4. Currently, my interest is to update a ReSTIR Project and take advantage of more robust ways to refit acceleration Structures and more powerful ways to access resources. I am working in Swift and have encountered a couple of puzzles: What is the 'accepted' way to create a MTL4BufferRange to store indices and vertices? How do I properly rewrite Swift code to build and compact an Acceleration Structure? I do realize that this is all in Beta and will happily look through Code Samples this Fall. If other guidance is available earlier, that would be fabulous! Thank you

I am developing a macOS terminal app, running on an M4 Pro, and using Metal. I am not able use float8 or float16, both reporting Variable has incomplete type 'float16' (aka '__Reserved_Name__Do_not_use_float16'). Based on the system I should be able to use these. Either it is because it is also compiling to Intel, which they are not allowed, or something else. Either way I have not been able to figure out how to get past this. IIs there a compiler setting I need to set to make this work? if so which one and what setting do I need? I only want to run this on M processes, on the latest version of OS so not interested in Intel version or backward compatibility.

Hello, I am experiencing an issue with programmatically capturing a GPU trace using MTLCaptureManager. The .gputrace file that is generated appears to be corrupted, and I'm looking for guidance or a solution. Description of the Problem: I am using MTLCaptureManager.sharedCaptureManager to capture a Metal frame and save it to disk. The generated .gputrace file is consistently reported as 0 bytes in size by the file system. Crucially, when I compress this 0-byte .gputrace file into a .zip archive, the resulting archive contains the full, expected data. After unzipping, the file can be opened and viewed correctly in Xcode. However,When inspecting the file's contents using NSFileManager in Objective-C (treating it as a directory), the internal structure is different from a .gputrace file captured directly from Xcode's Metal Debugger. capture in xcode capture in file Finally,When capturing multiple frames programmatically, the first captured frame contains valid buffer data. However, for subsequent frames (starting from the second frame), the corresponding buffer contents are all zero-filled. Frame 1: All MTLBuffer data is correctly captured and populated. Frame 2 and onward: The same MTLBuffer objects are present in the trace, but their contents are entirely 0 (i.e., the data is not captured or is corrupted). In this case, the on-screen display is normal, but the captured frame is incorrect. The frame captured directly in Xcode is also correct. Only the frame captured to a file is abnormal.

I mean…I want to use defaults rather than launching apps via open with the saved environment variables. This is pretty easy on iOS and other platforms. So what about in macOS?

Context I’m deploying large language models on iPhone using llama.cpp. A new iPhone Air (12 GB RAM) reports a Metal MTLDevice.recommendedMaxWorkingSetSize of 8,192 MB, and my attempt to load Llama-2-13B Q4_K (~7.32 GB weights) fails during model initialization. Environment Device: iPhone Air (12 GB RAM) iOS: 26 Xcode: 26.0.1 Build: Metal backend enabled llama.cpp App runs on device (not Simulator) What I’m seeing MTLCreateSystemDefaultDevice().recommendedMaxWorkingSetSize == 8192 MiB Loading Llama-2-13B Q4_K (7.32 GB) fails to complete. Logs indicate memory pressure / allocation issues consistent with the 8 GB working-set guidance. Smaller models (e.g., 7B/8B with similar quantization) load and run (8B Q4_K provide around 9 tokens/second decoding speed). Questions Is 8,192 MB an expected recommendedMaxWorkingSetSize on a 12 GB iPhone? What values should I expect on other 2025 devices including iPhone 17 (8 GB RAM) and iPhone 17 Pro (12 GB RAM) Is it strictly enforced by Metal allocations (heaps/buffers), or advisory for best performance/eviction behavior? Can a process practically exceed this for long-lived buffers without immediate Jetsam risk? Any guidance for LLM scenarios near the limit?

The following minimal snippet SEGFAULTS with SDK 26.0 and 26.1. Won't crash if I remove async from the enclosing function signature - but it's impractical in a real project. import Metal import MetalPerformanceShaders let SEED = UInt64(0x0) typealias T = Float16 /* Why ran in async context? Because global GPU object, and async makeMTLFunction, and async makeMTLComputePipelineState. Nevertheless, can trigger the bug without using global @MainActor let myGPU = MyGPU() */ @main struct CMDLine { static func main() async { let ptr = UnsafeMutablePointer<T>.allocate(capacity: 0) async let future: Void = randomFillOnGPU(ptr, count: 0) print("Main thread is playing around") await future print("Successfully reached the end.") } static func randomFillOnGPU(_ buf: UnsafeMutablePointer<T>, count destbufcount: Int) async { // let (device, queue) = await (myGPU.device, myGPU.commandqueue) let myGPU = MyGPU() let (device, queue) = (myGPU.device, myGPU.commandqueue) // Init MTLBuffer, async let makeFunction, makeComputePipelineState, etc. let tempDataType = MPSDataType.uInt32 let randfiller = MPSMatrixRandomMTGP32(device: device, destinationDataType: tempDataType, seed: Int(bitPattern:UInt(SEED))) print("randomFillOnGPU: successfully created MPSMatrixRandom.") // try await computePipelineState // ^ Crashes before this could return // Or in this minimal case, after randomFillOnGPU() returns // make encoder, set pso, dispatch, commit... } } actor MyGPU { let device : MTLDevice let commandqueue : MTLCommandQueue init() { guard let dev: MTLDevice = MPSGetPreferredDevice(.skipRemovable), let cq = dev.makeCommandQueue(), dev.supportsFamily(.apple6) || dev.supportsFamily(.mac2) else { print("Unable to get Metal Device! Exiting"); exit(EX_UNAVAILABLE) } print("Selected device: \(String(format: "%llX", dev.registryID))") self.device = dev self.commandqueue = cq print("myGPU: initialization complete.") } } See FB20916929. Apparently objc autorelease pool is releasing the wrong address during context switch (across suspension points). I wonder why such obvious case has not been caught before.

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!

Hi, I’m testing Unity’s Spaceship HDRP demo on iPhone 17 Pro Max and iPad Pro M4 (iOS 26.1). Everything renders correctly, and my custom MetalFX Spatial plugin initializes successfully — it briefly reports active scaling (e.g. 1434×660 → 2868×1320 at 50% scaling), then reverts to native rendering a few frames later. Setup: Xcode 16.1 (targeting iOS 18) Unity 2022.3.62f3 (HDRP) Metal backend Dynamic Resolution enabled in HDRP assets and cameras Relevant Xcode console excerpt: [MetalFXPlugin] MetalFX_Enable(True) called. [SpaceshipOptions] MetalFX enabled with HDRP dynamic resolution integration. [SpaceshipOptions] Disabled TAA for MetalFX Spatial. [SpaceshipOptions] Created runtime RenderTexture: 1434x660 [MetalFX] Spatial scaler created (1434x660 → 2868x1320). [MetalFX] Processed frame with scaler. [MetalFXPlugin] Sent RenderTexture (1434x660) to MetalFX. Output target 2868x1320. [SpaceshipOptions] MetalFX target set: 1434x660 [SpaceshipOptions] Camera targetTexture cleared after MetalFX handoff. It looks like HDRP clears the camera’s target texture right after MetalFX submits the frame, which causes it to revert to native rendering. Is there a recommended way to persist or rebind the MetalFX output texture when using HDRP on iOS? Unity doesn’t appear to support MetalFX in the Editor either: Thanks!

Hi all, I'm encountering an issue with Metal raytracing on my M5 MacBook Pro regarding Instance Acceleration Structure (IAS). Intersection tests suddenly stop working after a certain point in the sampling loop. Situation I implemented an offline GPU path tracer that runs the same kernel multiple times per pixel (sampleCount) using metal::raytracing. Intersection tests are performed using an IAS. Since this is an offline path tracer, geometries inside the IAS never changes across samples (no transforms or updates). As sampleCount increases, there comes a point where the number of intersections drops to zero, and remains zero for all subsequent samples. Here's a code sketch: let sampleCount: UInt16 = 1024 for sampleIndex: UInt16 in 0..<sampleCount { // ... do { let commandBuffer = commandQueue.makeCommandBuffer() // Dispatch the intersection kernel. await commandBuffer.completed() } do { let commandBuffer = commandQueue.makeCommandBuffer() // Use the intersection test results from the previous command buffer. await commandBuffer.completed() } // ... } kernel void intersectAlongRay( const metal::uint32_t threadIndex [[thread_position_in_grid]], // ... const metal::raytracing::instance_acceleration_structure accelerationStructure [[buffer(2)]], // ... ) { // ... const auto result = intersector.intersect(ray, accelerationStructure); switch (result.type) { case metal::raytracing::intersection_type::triangle: { // Write intersection result to device buffers. break; } default: break; } Observations Encoding both the intersection kernel and the subsequent result usage in the same command buffer does not resolve the problem. Switching from IAS to Primitive Acceleration Structure (PAS) fixes the problem. Rebuilding the IAS for each sample also resolves the issue. Intersections produce inconsistent results even though the IAS and rays are identical — Image 1 shows a hit, while Image 2 shows a miss. Questions Am I misusing IAS in some way ? Could this be a Metal bug ? Any guidance or confirmation would be greatly appreciated.

My app has a number of heterogeneous GPU workloads that all run concurrently. Some of these should be executed with the highest priority because the app’s responsiveness depends on them, while others are triggered by file imports and the like which should have a low priority. If this was running on the CPU I’d assign the former User Interactive QoS and the latter Utility QoS. Is there an equivalent to this for GPU work?

Hello, I am currently working on a research project under ENINCA Consulting, focused on advanced diagnostic tools for pseudorandom number generators (structural metrics, multi-seed stability, cross-architecture reproducibility, and complementary indicators to TestU01). To validate this diagnostic framework, I prototyped a small non-linear 64-bit PRNG (not as a goal in itself, but simply as a vehicle to test the methodology). During these evaluations, I observed something interesting on Apple Silicon (Mac M1): • bit-exact reproducibility between M1 ARM CPU and M1 Metal GPU, • full BigCrush pass on both CPU and Metal backends, • excellent p-values, • stable behaviour across multiple seeds and runs. This was not the intended objective, the goal was mainly to validate the diagnostic concepts, but these results raised some questions about deterministic compute behaviour in Metal. My question: Is there any official guidance on achieving (or expecting) deterministic RNG or compute behaviour across CPU ↔ Metal GPU on Apple Silicon? More specifically: • Are deterministic compute kernels expected or guaranteed on Metal for scientific workloads? • Are there recommended patterns or best practices to ensure reproducibility across GPU generations (M1 → M2 → M3 → M4)? • Are there known Metal features that can introduce non-determinism? I am not sharing the internal recurrence (this work is proprietary), but I can discuss the high-level diagnostic observations if helpful. Thank you for any insight, very interested in how the Metal engineering team views deterministic compute patterns on Apple Silicon. Pascal ENINCA Consulting

Hello, I am currently working on a research project under ENINCA Consulting, focused on advanced diagnostic tools for pseudorandom number generators (structural metrics, multi-seed stability, cross-architecture reproducibility, and complementary indicators to TestU01). To validate this diagnostic framework, I prototyped a small non-linear 64-bit PRNG (not as a goal in itself, but simply as a vehicle to test the methodology). During these evaluations, I observed something interesting on Apple Silicon (Mac M1): • bit-exact reproducibility between M1 ARM CPU and M1 Metal GPU, • full BigCrush pass on both CPU and Metal backends, • excellent p-values, • stable behaviour across multiple seeds and runs. This was not the intended objective, the goal was mainly to validate the diagnostic concepts, but these results raised some questions about deterministic compute behaviour in Metal. My question: Is there any official guidance on achieving (or expecting) deterministic RNG or compute behaviour across CPU ↔ Metal GPU on Apple Silicon? More specifically: • Are deterministic compute kernels expected or guaranteed on Metal for scientific workloads? • Are there recommended patterns or best practices to ensure reproducibility across GPU generations (M1 → M2 → M3 → M4)? • Are there known Metal features that can introduce non-determinism? I am not sharing the internal recurrence (this work is proprietary), but I can discuss the high-level diagnostic observations if helpful. Thank you for any insight, very interested in how the Metal engineering team views deterministic compute patterns on Apple Silicon. Pascal ENINCA Consulting

Hi, I am using xCode26.x. But my Metal4 classes are not compiling. I downloaded the sample code from Apple's website - https://developer.apple.com/documentation/Metal/processing-a-texture-in-a-compute-function. For example, I am getting errors like "Cannot find protocol declaration for 'MTL4CommandQueue'; I have hit a deadline. Any recommendations are very welcome. I have downloaded the Metal Tool chain. When I run the following commands on the terminal - xcodebuild -showComponent metalToolchain ; xcrun -f metal ; xcrun metal --version I get the following response - Asset Path: /System/Library/AssetsV2/com_apple_MobileAsset_MetalToolchain/86fbaf7b114a899754307896c0bfd52ffbf4fded.asset/AssetData Build Version: 17A321 Status: installed Toolchain Identifier: com.apple.dt.toolchain.Metal.32023 Toolchain Search Path: /Users/private/Library/Developer/DVTDownloads/MetalToolchain/mounts/86fbaf7b114a899754307896c0bfd52ffbf4fded /Users/private/Library/Developer/DVTDownloads/MetalToolchain/mounts/86fbaf7b114a899754307896c0bfd52ffbf4fded/Metal.xctoolchain/usr/bin/metal Apple metal version 32023.830 (metalfe-32023.830.2) Target: air64-apple-darwin24.6.0 Thread model: posix InstalledDir: /Users/private/Library/Developer/DVTDownloads/MetalToolchain/mounts/86fbaf7b114a899754307896c0bfd52ffbf4fded/Metal.xctoolchain/usr/metal/current/bin

I am puzzled by the setAddress(_:attributeStride:index:) of MTL4ArgumentTable. Can anyone please explain what the attributeStride parameter is for? The doc says that it is "The stride between attributes in the buffer." but why? Who uses this for what? On the C++ side in the shaders the stride is determined by the C++ type, as far as I know. What am I missing here? Thanks!

I have something like this drawing in an MTKView (see at bottom). I am finding it difficult to figure out when can the Swift-land resources used in making the MTLBuffer(s) be released? Below, for example, is it ok if args goes out of scope (or is otherwise deallocated) at point 1, 2, or 3? Or perhaps even earlier, as soon as argsBuffer has been created? I have been reading through various articles such as Setting resource storage modes Choosing a resource storage mode for Apple GPUs Copying data to a private resource but it's a lot to absorb and I haven't been really able to find an authoritative description of the required lifetime of the resources in CPU land. I should mention that this is Metal 4 code. In previous versions of Metal, the MTLCommandBuffer had the ability to add a completion handler to be called by the GPU after it has finished running the commands in the buffer but in Metal 4 there is no such thing (it it were even needed for the purpose I am interested in). Any advice and/or pointers to the definitive literature will be appreciated. guard let argsBuffer = device.makeBuffer(bytes: &args,... argumentTable.setAddress(argsBuffer.gpuAddress, ... encoder.setArgumentTable(argumentTable, stages: .vertex) // encode drawing renderEncoder.draw... ... encoder.endEncoding() // 1 commandBuffer.endCommandBuffer() // 2 commandQueue.waitForDrawable(drawable) commandQueue.commit([commandBuffer]) // 3 commandQueue.signalDrawable(drawable) drawable.present()

Is the pseudocode below thread-safe? Imagine that the Main thread sets the CAMetalLayer's drawableSize to a new size meanwhile the rendering thread is in the middle of rendering into an existing MTLDrawable which does still have the old size. Is the change of metalLayer.drawableSize thread-safe in the sense that I can present an old MTLDrawable which has a different resolution than the current value of metalLayer.drawableSize? I assume that setting the drawableSize property informs Metal that the next MTLDrawable offered by the CAMetalLayer should have the new size, right? Is it valid to assume that "metalLayer.drawableSize = newSize" and "metalLayer.nextDrawable()" are internally synchronized, so it cannot happen that metalLayer.nextDrawable() would produce e.g. a MTLDrawable with the old width but with the new height (or a completely invalid resolution due to potential race conditions)? func onWindowResized(newSize: CGSize) { // Called on the Main thread metalLayer.drawableSize = newSize } func onVsync(drawable: MTLDrawable) { // Called on a background rendering thread renderer.renderInto(drawable: drawable) }