The new version of the artificial intelligence inference performance measurement tool Procyon AI Inference Benchmark launched by UL Solutions this year is now also compatible with the macOS operating system. It mainly adds support for the Core ML AI inference engine to measure Mac machines equipped with Apple Silicon processors. artificial intelligence computing performance.
Through the Core ML AI inference engine, the computer vision benchmark test execution process of Procyon AI Inference Benchmark on the macOS operating system is the same as the artificial intelligence inference test workload on the Windows operating system environment. It can test CPU, CPU+GPU, Four types of computing combination performance, including CPU+artificial intelligence computing element and CPU+GPU+artificial intelligence computing element.
All test workloads are consistent with the current Mac model hardware environment and are fully compatible with GPU and artificial intelligence computing components. In addition to measuring the original hardware performance, the measurement process will also weigh the accuracy of artificial intelligence inference, which will compare FP16 (half-precision floating point operation performance) and INT precision inference results, making it easier to evaluate the measurement results. Performance performance.
On the other hand, the new version of Procyon AI Inference Benchmark already supports artificial intelligence inference engines including Microsoft WindowsML, Qualcomm SNPE, Intel OpenVINO, and NVIDIA TensorRT, and is compatible with Intel’s new 14th generation Core equipped with NPU design, codenamed Meteor Lake. series laptop processors, or Qualcomm’s new processors for Windows laptops.
The Procyon AI Inference Benchmark is currently open to the public for application. UL Solutions also stated that many industries have begun to use this tool to measure the artificial intelligence computing performance of notebook devices.
