Adobe adds NVIDIA optimizations, claiming up to twice the speed for creative tasks
The partnership also targets Frame.io’s cloud processing. A Nemotron-powered version of Brand Concierge is due for enterprise customers later in 2026.
Loading page…
The partnership also targets Frame.io’s cloud processing. A Nemotron-powered version of Brand Concierge is due for enterprise customers later in 2026.
Listen to this story
Adobe and NVIDIA expanded their March strategic partnership on October 15, 2026, targeting both GPU-intensive creative operations and the cloud pipeline that makes uploaded video searchable. Adobe claims up to 2× faster AI processing, editing, color, and effects on Photoshop, Premiere, and Substance 3D using RTX Spark capabilities, while Frame.io’s changes aim to speed search preparation. Its performance figures are early results from selected workloads, not platform-wide benchmarks. Brand Concierge’s Nemotron-powered enterprise version is planned for later in 2026, leaving broader availability and performance to be seen.
Photoshop’s GPU-accelerated compositing targets tools including Remove Tool, Gaussian Blur, and Light Leaks.
Premiere optimizations include playback, rendering, effects, color, Enhance Speech, and Object Mask; Substance 3D Painter gets faster texturing.
Adobe reports Frame.io embedding generation about 5× faster, P95 processing latency 4–5× better, and search proxies about 80% smaller.
Adobe and NVIDIA are putting more image editing, video production and media-search processing onto NVIDIA computing technology. Their October 15, 2026 partnership expansion covers Photoshop, Premiere, Substance 3D, Frame.io and Brand Concierge, with changes aimed at speeding both creative work and the preparation that makes uploaded media searchable.
The agreement builds on a strategic partnership announced in March. It combines Adobe’s creative and marketing workflows with NVIDIA’s open models, software libraries, research and accelerated computing.
For Photoshop, Premiere and Substance 3D, the hardware target is NVIDIA’s RTX Spark superchip. Adobe says the applications use its unified memory and AI computing capabilities to deliver up to twice the speed for AI processing, editing, color and effects—not a uniform increase for every operation.
Frame.io’s changes happen in the cloud. Its updated pipeline uses NVIDIA NVDEC for hardware-accelerated video decoding, and TensorRT with CUDA for content preprocessing and embedding generation. Those embeddings are representations used in content discovery. Adobe says moving these computing-heavy steps onto NVIDIA GPUs makes uploaded media searchable sooner and helps larger libraries refresh more efficiently.
The enterprise conversational work uses NVIDIA’s Nemotron open models. Adobe and NVIDIA have optimized Brand Concierge with those models, but Adobe says this version will reach enterprise customers later in 2026.
NVIDIA has also selected Brand Concierge for personalized conversations on NVIDIA.com. Adobe says those interactions will draw on organizational data and content, using Nemotron to tailor responses to customer needs. NVIDIA is therefore a planned user as well as the technology partner.
Loading discussion...
Join the conversation
Explain which would make a bigger difference to your creative work.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.