The most recent version of Primate Labs’ cross-platform benchmarking programme, Geekbench 6, has been made available. This was released three years after Geekbench 5’s debut.
The updated version comes with updated and better workloads for assessing CPU and GPU performance. Additionally, it now evaluates effectiveness in fresh application domains like augmented reality and machine intelligence.
Geekbench 6: Features
Geekbench’s latest version includes several new tests, including workloads that:
- Background blur in videoconferencing streams.
- Automatically remove unwanted objects from photos.
- Using machine learning models, detect and tag objects in photos.
- Scripting languages are used to analyse, process, and convert text.
Geekbench 6’s multi-core benchmark tests have also been updated. Instead of assigning tasks to each core, the tests now assess how well the cores collaborate. This method improves multi-core tests and more accurately measures heterogeneous core performance.
This feature is useful for desktops and laptops, which increasingly use “performance” and “efficient” cores. Furthermore, the new Geekbench 6 includes Vulkan support, the next-generation cross-platform graphics and compute API.
The company has also updated the datasets that the workloads process to better align with today’s file types and sizes. This includes the following:
- images with higher resolution in image tests.
- In navigation tests, larger maps are used.
- In the PDF and HTML5 tests, larger, more complex documents were used.
- In the developer tests, there are more (and larger) files.
For Android, iOS, Windows, macOS, and Linux users, Geekbench 6 is now available for download. To celebrate the release of the newer version, Geekbench is offering 20% off Geekbench 6 Pro licences until February 28. It will remain free to use for personal purposes.
Qualcomm Technologies stated regarding the Geekbench 6 version: