US' Frontier.cool launches FPGA-based MPSoC for textile industry

The Fabric Meta Chip is a ‘dream-team’ collaboration between Frontier.cool and E-Elements — the single-chip solutions experts. What's more, the Fabric Meta MPSoC uses Xilinx Zynq FPGAs with arm processor architecture. The Fabric Meta MPSoC was developed to convert physical fabric into digital material, thus allowing virtual apparel to be designed directly from real-world textile fabrics — expanding the benefits of virtual fashion and powering up the textile supply chain, the company said in a media statement.

This innovative world-first can be installed in textile and garment factories and fabric manufacturers to speed up digital fabric layer computation, facilitating the connection to 3D apparel design software and enabling the representation of the physical fabric in the virtual apparel or fashion space in real-time. Frontier realised a critical need for robust computing power for edge computing to give enhanced power and higher resolution — hence the birth of the Fabric Meta MPSoC.

“In the future, personalised virtual clothes worn by digital doppelgangers in the metaverse could be faithfully replicated in the real world via AI fabric computing — from the cloud to the edge. Thanks to this seismic shift, the product development process can now be achieved in the virtual world, regardless of geographical location or time limitations,” the founder of Frontier.cool Victor Chao said in a statement.

The evolution of the semiconductor process to integrate heterogeneous multi-core processing units in a single chip can provide high-speed computing acceleration for various applications. In addition, AI has several diverse features that are ideal for using FPGA to provide a customised design for specific applications and to increase the added value of the entire system. Since the fashion metaverse requires digital twins to represent virtual clothing, and virtual clothing requires digital fabrics that can be connected to 3D design software, the Fabric Meta MPSoC, which uses AI to convert physical materials into digital twins, helps the textile supply chain and virtual fashion to work together in harmony. 

Frontier.cool’s AI engine was developed by advanced machine learning technology to analyse 2D digital fabrics layer-by-layer and quickly generate 3D fabric layers that can be stored and downloaded as, in as little as three minutes, high fidelity 3D and 4D-ready U3M files. The texture of the 3D material is high quality, high resolution, clear, and realistic. In fact, it is very difficult to tell the difference between the original material and the AI-generated material.

Material digitisation tasks can now be carried out and scaled without the usual costly challenges and with the support of Frontier.cool's AI and machine learning technology, consistent and accurate results are delivered each and every time. Frontier.cool's unique process enables its users to use a mid-range scanner to scan and upload materials in as little as 30 seconds.

Fibre2Fashion News Desk (GK)