Primate Labs has unveiled Geekbench 7, the latest iteration of its renowned benchmarking suite designed to assess CPU and GPU performance across various devices. This release introduces significant enhancements aimed at providing more accurate and relevant performance metrics.
Refined Multi-Core Benchmarking
Geekbench 7 features a redesigned multi-core benchmark that better mirrors real-world application behavior. Unlike previous versions, where workloads were uniformly multi-threaded, the new approach ensures that tasks are executed in multi-threaded mode only if they typically run that way in actual applications. This refinement offers users a multi-core score that more accurately reflects a device’s performance during everyday tasks.
Advanced GPU Testing
The GPU benchmark has been overhauled to focus on contemporary uses such as machine learning and content creation. New tests include real-time video filter application, image upscaling using machine learning, and background blurring in video conferencing. Additional GPU tasks encompass RAW image processing, LUT-based video color grading, path tracing, and fluid simulation. Notably, support for Nvidia’s CUDA has been reinstated alongside existing APIs like OpenCL, Vulkan, and Metal.
Expanded Workloads and Media Testing
To better represent modern computing demands, Geekbench 7 incorporates more complex datasets. File Compression and PDF Viewer workloads now feature greater variability, while developer and image processing tests include a broader range of assets and image formats. New media workloads assess audio and video encoding, decoding, and processing capabilities, with tests like Whisper speech recognition and AV1 and Opus encoding simulating scenarios such as video conferencing and content consumption.
Gaming and Photo Editing Enhancements
For gaming enthusiasts, a new Game Physics workload utilizes the Jolt Physics engine, commonly employed in popular video games, to evaluate physics processing performance. The Photo Editor test has been expanded to encompass a richer set of real-world edits, and the Photo Library workload now supports importing and processing image formats like JPEG XL and DNG.
Geekbench 7 is available for iOS, macOS, and other platforms. The application remains free for personal use, with a Pro version offered at a 20% discount off the regular $99 price through August 6.
With these updates, Geekbench 7 provides a more comprehensive and realistic assessment of device performance, aligning closely with current computing trends and user needs. This evolution is particularly relevant as devices increasingly handle complex tasks like machine learning and advanced content creation, making accurate performance metrics essential for both consumers and professionals.