Primate Labs has officially released Geekbench 7, the latest version of its cross-platform benchmarking tool for CPUs and GPUs.
The CPU benchmark now includes new media-related workloads that test tasks such as AV1 screen-sharing video encoding, Opus audio compression, and video playback with live captions using the Whisper speech recognition model.
These workloads are designed to reflect common activities like video calls, podcasts, voice memos, and content streaming.
Other additions include a new game physics test based on the Jolt Physics engine, expanded photo editing tasks, and support for newer image formats such as JPEG XL and DNG in the photo library workload.
Primate Labs also redesigned the multi-core benchmark. Instead of running every test across multiple CPU cores, Geekbench 7 only uses multi-threading for workloads that typically do so in real-world applications. The company says this provides a more realistic measurement of multi-core performance.
The GPU benchmark has also been updated with a stronger focus on machine learning and content creation. New tests include face tracking with real-time filters, AI image upscaling, video background blur, RAW image processing, video color grading, path tracing, and fluid simulation.
Geekbench 7 also adds CUDA support alongside OpenCL, Vulkan, and Metal, allowing NVIDIA GPU users to benchmark using CUDA.
The benchmark uses larger and more varied datasets than Geekbench 6, including additional file archives, PDF documents, and image assets to represent modern workloads better.
Geekbench 7 is free for personal use and is available for download starting today on Android, iOS, Windows, macOS, and Linux.
ManilaShaker is a tech media producing insightful and helpful content for our local and growing international audience. Our goal is to create a premier Philippine digital consumer electronics resource that provides the most objective reviews and comparisons globally.
