Nvidia Studio with full artificial intelligence: focus on Adobe Super Resolution and DaVinci Neural Engine

Nvidia is not just gaming but also pay attention to the professionals who use its GPUs, more precisely the technologies made available, to produce content and not just passively use it. When it comes to RT core and Tensor core within the latest desktop and mobile GPUs, not everyone knows that those units are not only used to manage the graphic novelties of video games, but also for accelerate productivity.

The US company, under the brand Nvidia Studio, is constantly working to improve the relationship between its GPUs and the creative software, which they have integrated as never before in the last two years artificial intelligence functions to accelerate once-time-consuming and resource-intensive tasks. In fact, it is estimated that 80% of creative activities are repetitive and distract the content creator from creating new content.

With the release of the latest 461.92, Nvidia took stock of the Announcements affecting the world of content creator. On the subject of artificial intelligence, Adobe has integrated Super Resolution on Camera Raw, announcing its arrival in the summer in Lightroom and Lightroom Classic: this is a function that enlarges photos while keeping edges clean and preserving important details. With the click of a button, Super Resolution enlarges and enhances a 10 megapixel photo into a 40 megapixel image ready to be printed.

Bicubic resampling (left), Super Resolution (right)

According to Nvidia, a laptop with RTX 3070 is up to 27 times faster in using Super Resolution compared to a similar product with a CPU with integrated graphics. Another Adobe Lightroom feature, Enhance Images, uses machine learning to sharpen and improve image quality, up to five times faster on Nvidia GPUs.

Regarding Adobe Photoshop, there are some real-time AI novelties like i neural filters: This is a set of features that allows you to make complex changes to images in seconds. With Smart portrait In Photoshop, you can make changes to facial features, such as gaze direction and lighting angles, simply by dragging a slider. Done manually, these adjustments would take hours of tedious work. Artificial intelligence allows you to do them almost instantly.

Another optimization concerns the DaVinci Neural Engine di DaVinci Resolve 17, enhanced and joined by new tools such as Magic Mask to easily select and remove people from backgrounds and automatically create masks for an entire person or specific features such as the face or arms.

Smart Reframe quickly changes 16: 9 HD or Ultra HD footage into square or vertical versions automatically detecting and tracking objects and keeping them in the frame, replacing the need to do it manually. Scene Cut Detection also analyzes the content on the timeline and applies edits to isolate each scene, instead of applying the edits individually.

To spend less time in the editing room, Adobe Premiere Pro’s Scene Edit Detection feature, recently introduced in the Premiere Pro public beta, leverages Nvidia GPUs to quickly detect and label cuts in final video files, eliminating lengthy scrubbing sessions. to make manual changes. And once it’s time to export, the hardware encoder built into the GeForce GPUs exports high-resolution video up to 5x faster than the CPU alone.

The company’s graphics chips also support the denoising based on Blender’s artificial intelligence– This is a process that anticipates the final images from the partially rendered results when designing in the viewport.

Finally, among the announcements this month, Nvidia recalled the arrival on the market of Asus ZenBook Pro Duo 15 OLED, a 15-inch notebook (4K HDR) equipped with a second tilting ScreenPad Plus screen. This laptop is based on the Intel platform up to the Core i9 coupled with a mobile GPU up to the GeForce RTX 3070, a PCIe 3.0 x4 SSD up to 1 TB and a maximum of 32 GB of DDR4 memory. It is an Nvidia Studio certified product, so it is particularly suitable for content creators.