EnCharge AI has introduced the EN100, an AI accelerator chip that utilizes analog in-memory computing technology to deliver high-performance compute capabilities on-device. This chip is engineered to provide over 200 TOPS (trillion operations per second) of compute power while operating within the power constraints typical of edge and client devices. Its architecture is designed for efficiency, enabling advanced AI inferences to be performed locally on devices such as laptops without the need for extensive cloud-based processing or additional hardware accelerators.

The EN100 is offered in two distinct form factors to cater to different deployment environments. The M.2 version is optimized for portable devices, delivering robust AI performance within an 8.25W power envelope to support applications on lightweight systems. In contrast, the PCIe version is intended for workstation use, featuring four NPUs that collectively aim for compute capacities near 1 PetaOPS, making it suitable for demanding professional applications. This dual form-factor strategy positions the EN100 as a versatile solution for enhancing on-device AI capabilities across a range of platforms while addressing both performance and energy efficiency requirements.

Image Credit: Encharge