Metagenomi, a California-based biotechnology company, is leveraging Amazon Web Services’ custom AI chips to accelerate the discovery of new gene-editing tools. The partnership uses AWS’s Inferentia processors to power Metagenomi’s AI models, which analyze vast microbial datasets to identify and optimize enzymes capable of precisely editing DNA. Traditionally, such large-scale computations relied on expensive GPU infrastructure. By shifting to Amazon’s specialized hardware, Metagenomi reports up to a 56% reduction in computational costs and the ability to generate millions of enzyme candidates more efficiently.
This collaboration highlights a growing convergence between biotechnology and cloud computing. Generative AI, once focused mainly on text or image creation, is now being applied to protein and enzyme design, potentially transforming the pace of therapeutic innovation. Metagenomi’s expanding enzyme library could lead to more versatile and targeted treatments for genetic diseases, while demonstrating how AI-driven discovery can reduce barriers to entry in biotech research. For Amazon, the project showcases its custom silicon’s competitiveness against established GPU systems, opening new markets in the life sciences. Though still early-stage, the work underscores how advances in computing hardware are becoming foundational to next-generation medicine and gene-editing breakthroughs.
Metagenomi taps Amazon's custom chips to develop gene-editing tools
Biotech firm Metagenomi said it is using artificial intelligence chips from Amazon.com's cloud computing unit to power some of its work in developing gene-editing technologies.
www.reuters.com