Global AI benchmark organization MLCommons releases MLPerf Client, a benchmark application that measures the AI performance of PCs



In environments where privacy and security are important, local AI processing on your own device may be more suitable than using online AI services like ChatGPT. However, running AI locally requires a PC with a certain level of performance.

MLCommons , an AI industry group, recently released a new AI benchmark app for PCs called MLPerf Client on Wednesday, July 30, 2025.

MLPerf Client Benchmark
https://mlcommons.org/benchmarks/client/

MLCommons Releases MLPerf Client v1.0: A New Standard for AI PC and Client LLM Benchmarking - MLCommons
https://mlcommons.org/2025/07/mlperf-client-v1-0/

MLPerf Client is a benchmark app that measures the performance of large-scale language models, such as Llama 2 7B Chat, Llama 3.1 8B Instruct, Phi 3.5 Mini Instruct, and Phi 4 Reasoning 14B, when run on a PC. The benchmark measures the performance of operations using AI models, such as code analysis, content generation, creative writing, and summarization.



MLPerf Client can be run on a PC that meets the following specifications: Windows 11 and macOS 15.5 are supported, and at least 400GB of free storage space is required to run all benchmarks.
- A PC equipped with an AMD Radeon RX 7000 series or RX 9000 series with 8GB or more of VRAM
・PC equipped with AMD Ryzen AI 9 series
・PC equipped with Intel Arc B series with 10GB or more of VRAM
- A PC equipped with an Intel Core Ultra Series 2 (Lunar Lake, Arrow Lake) with 16GB or more of RAM
- A PC equipped with an NVIDIA GeForce RTX 2000 series with 8GB or more of VRAM
- A PC equipped with a Qualcomm Snapdragon X Elite with 32GB or more RAM

MLPerf Client can be downloaded for free from the link below.

Releases · mlcommons/mlperf_client
https://github.com/mlcommons/mlperf_client/releases



The source code for MLPerf Client is also available at the following link:

GitHub - mlcommons/mlperf_client: MLPerf Client is a benchmark for Windows and macOS, focusing on client form factors in ML inference scenarios.
https://github.com/mlcommons/mlperf_client/



in Software, Posted by log1o_hf