Samsung’s new Exynos 2600 SoC appears in die‑shots released by TechPowerUp. Built on the company’s in‑house 2 nm GAAFET node, the monolithic chip packs a 3‑tier CPU, an AMD‑licensed RDNA 4 iGPU, and a powerful NPU, positioning Samsung close to rivals on transistor density and performance today market.

The CPU follows a 1X+3P+6E core layout. A single C1‑Ultra core tops 3.80 GHz with 3 MB L2 cache, while three C1‑Pro cores hit 3.25 GHz and six efficiency cores run at 2.75 GHz. All ten cores share 16 MB of L3 cache, giving the chip balanced peak and low‑power capability for high‑end smartphones and ultraportables through efficient task switching.

The Xclipse 960 iGPU, sourced from AMD’s RDNA 4, hosts 16 compute units and 1,024 stream processors, plus 64 TMUs and 32 ROPs, all backed by 4 MB L2 cache. Its LPDDR5X interface, USB 3.2 PHYs, and eDP support position the SoC for modern display and connectivity demands in next‑gen devices for high‑end gaming applications and enterprise workstations with seamless integration.

Samsung’s NPU delivers over 100 % more throughput than the Exynos 2500, reaching 59 TOPS with a 32K MAC core array and 8 MB scratchpad RAM. Coupled with image processors, DSPs, and on‑chip NV storage, the Exynos 2600 aims to consolidate performance, AI, and power efficiency, threatening Qualcomm’s dominance in flagship smartphones for global market users.