| Offset | Size (bytes) | Description | |--------|--------------|-------------| | 0x00 | 4 | Magic number ( MACE or MCLP ) | | 0x04 | 4 | Version (e.g., 0x00010000) | | 0x08 | 4 | Number of kernels | | 0x0C | 4 | Total binary size | | 0x10 | 4 | Offset to kernel table | | ... | ... | Vendor-specific headers (Movidius blob) | | ... | ... | Compiled ELF-like sections (VPU instructions) | | End | ... | Possibly signature/hash |
Stores pre-compiled OpenCL kernels for fast GPU inference. Context: Used by the Xiaomi MACE deep learning framework.
Directly in the root of the internal storage /storage/emulated/0/ if an older or poorly configured app version generated it. Is mace-cl-compiled-program.bin Safe?
that loads the binary via MACE runtime.
Knowing this can help determine exactly and how to keep it from growing . Share public link
: Especially on Xiaomi, Redmi, POCO, and Black Shark devices, where MACE powers AI scene detection, night mode, and portrait blur.
#include "mace/public/mace.h" // Initialize GPU context with precompiled paths DeviceType device_type = DeviceType::GPU; mace::MaceEngineConfig config(device_type); std::shared_ptr gpu_context = mace::GPUContextBuilder() .SetStoragePath("/data/local/tmp/mace_run/cache/") .SetOpenCLBinaryPaths("path/to/mace-cl-compiled-program.bin") .SetOpenCLParameterPath("path/to/mace-cl-tuned-parameter.bin") .Finalize(); config.SetGPUContext(gpu_context); Use code with caution. Troubleshooting Common Errors mace-cl-compiled-program.bin
It is worth clarifying that mace-cl-compiled-program.bin is distinct from other concepts with similar names. For example, there is "MACE," an abbreviation for a machine-learning interatomic potential for chemistry, which is unrelated to Xiaomi's MACE. Similarly, there is an unrelated MACE (Miami Area Control Engine) audio codec for FFmpeg, which has nothing to do with Xiaomi's MACE or OpenCL.
The file is a specialized binary cache file generated by the Mobile AI Compute Engine (MACE) , a deep learning inference framework developed by Xiaomi to optimize neural network performance on mobile and edge devices. When MACE executes a machine learning model using a device's Graphics Processing Unit (GPU), it utilizes the OpenCL (Open Computing Language) framework. The mace-cl-compiled-program.bin file stores the precompiled OpenCL device kernels, effectively saving the processed code so the application does not have to compile the shaders from scratch during subsequent launches.
# Step 1: Write MACE YAML config (mace.yml) models: my_model: platform: myriad model_file: model.tflite weight_file: model.data input_nodes: input output_nodes: output runtime: opencl | Offset | Size (bytes) | Description |
For automated object removal, video stabilization, and smart cropping. Frequently Asked Questions Is mace-cl-compiled-program.bin a virus or malware?
Any app downloaded from the Google Play Store that integrates MACE for on-device AI processing can create this file, regardless of your phone’s manufacturer. Where is the File Located?
Applications using MACE (like camera apps, AI editors) can load models quicker, leading to better user experiences. How mace-cl-compiled-program.bin Works Context: Used by the Xiaomi MACE deep learning framework