Geekbench is one of the best tools for comparing one computer, phone, or tablet against another in raw performance. There’s now a new test you can try out: Geekbench AI.
Geekbench includes a CPU benchmark test and GPU test, and there was a separate “Geekbench ML” test for evaluating machine learning performance. Geekbench has released a replacement for the ML benchmark, called Geekbench AI, which is intended to more directly measure how a device can handle a wider set of AI-centric workloads.
The company said in a blog post, “Geekbench AI is a benchmarking suite with a testing methodology for machine learning, deep learning, and AI-centric workloads, all with the same cross-platform utility and real-world workload reflection that our benchmarks are well-known for. Software developers can use it to ensure a consistent experience for their apps across platforms, hardware engineers can use it to measure architectural improvements, and everyone can use it to measure and troubleshoot device performance with a suite of tasks based on how devices actually use AI.”
Geekbench AI tests your device’s performance with common AI tasks and software frameworks, like OpenVINO on Linux and Windows and the Core ML Neural Engine on Apple devices, and combines the results into three numbers. You get a “Full Precision Score,” a “Half Precision Score,” and a “Quantized Score,” reflecting performance in tasks that rely on exact output as well as tasks with less specific results.
Just like the normal Geekbench tests, Geekbench AI can submit results to the website for easy comparisons across different devices. The new Geekbench AI results page shows the M4 iPad Pro at the top of the charts (identified as “iPad16,5”), followed by the Samsung Galaxy S24 series. The Asus Zenfone 11 Ultra (listed as “ASUS_AI2401_H”) also shows up alongside other flagship Android phones, mixed in with newer PC hardware like the NVIDIA GeForce RTX 4090 and Ryzen 9 9950X.
You can download Geekbench AI 1.0 starting today on Android, iPhone and iPad, and desktop platforms.
Source: Geekbench
