ARM Architecture

Flex Logix Broadcasts Manufacturing Availability of InferX X1M Boards for Edge AI Imaginative and prescient Techniques

X1M Boards Ship Excessive Efficiency and Extremely Low Energy in a Tiny M.2 Type Issue

MOUNTAIN VIEW, Calif., March 30, 2022— Flex Logix® Applied sciences, Inc., provider of the most-efficient AI edge inference accelerator and the main provider of eFPGA IP, in the present day introduced manufacturing availability of its InferX™ X1M boards. At roughly the scale of a stick of gum, the brand new InferX X1M boards pack excessive efficiency inference capabilities right into a low-power M.2 type issue for area and energy constrained purposes reminiscent of robotic imaginative and prescient, industrial, safety, and retail analytics.

“With the overall availability of our X1M board, clients designing edge servers and industrial imaginative and prescient methods can now incorporate superior AI inference capabilities with high-accuracy, excessive throughput and low energy on advanced fashions,” stated Dana McCarty, Vice President of Gross sales and Advertising and marketing for Flex Logix’s Inference Merchandise. “By incorporating an X1M board, clients cannot solely design new and thrilling new AI capabilities into their methods, however additionally they have a quicker path to manufacturing ramp versus designing their very own customized card design.”

Flex Logix InferX™ X1M board

Concerning the InferX X1M Board

That includes Flex Logix’s InferX X1 edge inference accelerator, the InferX X1M board provides probably the most environment friendly AI inference acceleration for superior edge AI workloads reminiscent of Yolov5. The boards are optimized for giant fashions and megapixel pictures at batch=1. This offers clients with the high-performance, low-power object detection and different high-resolution picture processing capabilities wanted for edge servers and industrial imaginative and prescient methods.

The InferX X1M M.2 board suits inside the low energy necessities of the M.2 specification. To assist its clients to market rapidly, Flex Logix additionally offers a set of software program instruments to accompany the boards. This consists of instruments to port educated ONNX fashions to run on the X1M, and easy runtime framework to help inference processing inside each Linux and Home windows.

Additionally included within the software program instruments is an InferX X1 driver with exterior APIs designed for purposes to simply configure & deploy fashions, in addition to inside APIs for dealing with low-level features designed to regulate and monitor the X1M board.

About Flex Logix

Flex Logix is a reconfigurable computing firm offering AI inference and eFPGA options primarily based on software program, methods and silicon. Its InferX X1 is the business’s most-efficient AI edge inference accelerator that can carry AI to the plenty in high-volume purposes by offering a lot increased inference throughput per greenback and per watt. Flex Logix’s eFPGA platform permits chips to flexibly deal with altering protocols, requirements, algorithms, and buyer wants and to implement reconfigurable accelerators that pace key workloads 30-100x in comparison with processors. Flex Logix is headquartered in Mountain View, California and in addition has places of work in Austin, Texas. For extra info, go to

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button