DeepX's DX-M2 Edge AI Chip Receives CES 2027 Innovation Award
The AMW Read
The award and tape-out update a known edge-chip roadmap, while performance and customer adoption remain unverified.
DeepX's DX-M2 Edge AI Chip Receives CES 2027 Innovation Award
DeepX said its DX-M2 generative AI chip received a CES 2027 Innovation Award in the artificial intelligence category. The chip is designed to run large language models and vision-language models directly on robots and industrial devices. It uses Samsung Foundry's SF2P 2-nanometer process and targets inference at approximately 5 watts. DeepX completed mask tape-out in early November 2026, with first silicon planned for March 2027.
The design extends DeepX's earlier push to position low-power chips for physical AI. On-device language and vision-language inference could reduce dependence on cloud services and expand what robots and industrial systems can process locally. The 5-watt figure remains a development target, however, and the award does not establish the chip's performance in customer deployments.
Builders and investors should watch the first silicon and customer evaluations for evidence that DX-M2 can meet its power target while running useful workloads. DeepX plans to provide evaluation samples in the second quarter of 2027, begin initial production shipments in the fourth quarter, and expand mass production in 2028. Its earlier DX-M1 offers some commercial context: DeepX reported 88 purchase orders worth approximately $13.3 million across more than 10 countries as of October 2026. Those orders concern the first-generation chip, so they do not establish demand for DX-M2.



