Login to Continue Learning
Intel has demonstrated that its 18A process technology is ready for System-on-Chips (SoCs) not using the x86 architecture, signaling a broader customer adoption. Initially, Intel plans to test the node with products like Panther Lake. Its success will determine whether customers will adopt this process.
In a recent live demo, Intel showcased a non-x86 chip utilizing the 18A process. While the specific IP technology used in the test was not disclosed, this indicates that the division aims to become a world-class contract manufacturer. During the demo, a reference SoC ran different workloads such as 3D gaming, animation, and 4K video streaming. The chip featured seven CPU cores from various performance classes, with additional PCIe and controller IP sourced externally.
The live workload demo highlighted core specialization and how the 18A node can handle diverse real-world tasks efficiently. Intel’s developer tools, including VTune Profiler, are now optimized to support non-x86 SoCs, enhancing CPU utilization.
Intel aims for the 18A process to tap into major fabless companies like Apple, Qualcomm, and NVIDIA, making it a competitive alternative to TSMC N2 in both x86 and broader markets. This development opens up prospects for integration into Apple’s M-series, Qualcomm’s Snapdragon lineup, and even NVIDIA’s Grace CPU series.
The key to the 18A’s success is its ability to attract major fabless companies. By being deployable on non-x86 chips, it opens up opportunities for integration into Apple’s M-series, Qualcomm’s Snapdragon lineup, and even NVIDIA’s Grace CPU series. This also means that 18A can compete with TSMC N2 in a much broader market, not just the x86 segment.
