Efabless Unveils Custom Silicon Platform for Edge ML
Efabless has launched a custom chip platform that enables customers to develop tailor-made machine learning chips in small batches, enhancing efficiency for edge ML applications.
Published on April 14, 2025
Efabless has introduced its new chipIgnite ML platform, a custom silicon solution engineered for edge machine learning applications. The platform enables customers to develop tailored ML chips in batch sizes as small as 100, offering enhanced performance and power efficiency. Leveraging pre-hardware simulation for profiling and optimization of ML inference workloads, chipIgnite ML aims to reduce development time and improve deployment in resource-constrained environments.
In addition to the flexible production runs, Efabless’s platform integrates SensiML’s open-source AutoML development tools, providing a streamlined approach to design and prototyping. The design kit for early evaluation was made available in November 2024, with the first shuttle for prototyping scheduled for April 2025, marking a significant leap in custom edge AI innovation as reported by multiple sources since early October 2024.