Samsung announced the Exynos 2600 last week as the world’s first 2nm chipset. It is expected to make its debut on the Galaxy S26 series in Q1, 2026, and we now have some additional details regarding the 2600’s AI capabilities.
The new chip is equipped with an all-new NPU, which is claimed to offer a 113% improvement over the Exynos 2500, and we now get confirmation that it will also get a significant boost for on-device AI tasks. Korean AI firm NotaAI revealed that its AI model optimization tech will be used on the Exynos 2600 to run larger AI models directly on-device.
🤝 Thrilled to share that Nota AI’s AI model optimization technology will be supplied for Samsung Semiconductor’ next-generation AP, Exynos 2600. #NotaAI #SamsungSemiconductor pic.twitter.com/BZkHFQXXft
— Nota AI (@nota_ai) December 30, 2025
NotaAI’s NetsPresso AI optimization platform allows for a substantial reduction in the size of AI models while still offering accurate results. The firm specializes in optimizing device-tailored AI models, which would allow Samsung engineers to run larger AI models on-device without needing Internet access.
The two companies previously worked together on the Exynos 2500 and 2400 AI optimizations and co-developed Samsung’s Exynos AI Studio toolchain.