Tech Desk 
Aug 16, 2024
Geekbench, a popular benchmarking utility, recently released a new cross-platform tool to examine the performance of devices under artificial intelligence (AI)-heavy workloads.
Geekbench AI measures a device’s Central processing unit (CPU), Graphics processing unit (GPU), and Neural Processing Unit (NPU) to determine how well it can manage machine learning applications.
Geekbench developer Primate Labs has been developing the software utilising the name Geekbench ML, which released in preview in 2021, but transferred the name to AI for unidentified reasons.
Read more: Huawei develops Ascend 910C AI chip to rival Nvidia's dominance — Report
To find how different hardware responds to various AI-related tasks, it examines performance based on both accuracy and speed, with support for multiple frameworks, including Open Neural Network Exchange (ONNX), CoreML, TensorFlow Lite, and OpenVino.
It delivers three scores that are full precision, half precision, and quantised. Primate Labs said the scores also have an accuracy measurement to estimate how close a workload’s outputs are to the truth, “or how accurately that model can do what it’s supposed to do.”
"We’ll need more time with devices running on-device AI such as Copilot Plus PCs and all of the new phones to see how performance in real-world tasks correlates with Geekbench AI’s numbers. Checking frame rates or loading times is one thing — now we might be checking the accuracy of predictive text or what a generative AI-enabled image editor comes up with," Primate Labs added.
