Photoshop for Apple Silicon & Image Upscaling - Adobe

Adobe Optimizes Photoshop for Apple Silicon and Introduces Super Resolution

Adobe is rapidly adapting its imaging software to fully utilize Apple’s new processors, beginning with the M1-based MacBook Pro and MacBook Air. Following the release of native Lightroom and Camera Raw versions, an Apple Silicon-optimized version of Photoshop is now available.

Performance Gains with Apple Silicon

This new version delivers substantial performance improvements compared to the Intel version running through Rosetta 2 emulation. Internal testing indicates users can expect performance increases of up to 1.5x across various Photoshop features.

Adobe intends to continue optimizing the software for Apple Silicon in collaboration with Apple, promising further enhancements over time. Currently, some features, such as “Invite to Edit Cloud Documents” and “Preset Syncing,” are not yet available in the M1-compatible version, but will be included in future updates.

Introducing Super Resolution

Alongside the Apple Silicon Photoshop update, Adobe is launching a new Super Resolution feature within the Camera Raw plugin, with a future release planned for Lightroom. This feature leverages machine learning to enlarge images while maintaining a high level of detail.

Previously, Adobe offered a super resolution option that combined multiple exposures. However, this new implementation works effectively with a single image.

How Super Resolution Works

This technology effectively brings the “Computer, enhance” concept from science fiction to reality. It builds upon Adobe’s prior work with the “Enhance details” feature in Photoshop. Similar functionality can be found in Pixelmator Pro’s “ML Super Resolution” feature, though it utilizes a different machine learning model and dataset.

Adobe’s Super Resolution outputs an image with double the horizontal and vertical resolution, resulting in a fourfold increase in total pixels. This process preserves detail and sharpness, enabling larger prints from images that previously lacked the necessary resolution.

The feature also proves beneficial for cropping photos, allowing for tighter shots without sacrificing image quality.

Hardware Acceleration

Super Resolution benefits significantly from GPUs optimized for machine learning tasks, including CoreML and Windows ML. Apple’s M1 chip, with its dedicated Neural Engine, is particularly well-suited for this purpose. Similarly, Nvidia’s RTX series GPUs and their TensorCores are also highly capable.

Photoshop for iPad Updates

Adobe has also released significant updates for Photoshop for iPad. These include version history for Cloud Documents, enabling non-local storage. Users can now also store versions of Cloud Documents offline and edit them locally on their devices.