(London, September 2, 2024) Google DeepMind, in collaboration with the Materials Laboratory at the University of Cambridge, has released a revolutionary AI material design platform called ForgeSteel. The platform successfully developed a new gradient structure mold steel precision plate by simulating microstructure evolution through generative adversarial networks (GANs). Its heat and fatigue resistance is three times higher than that of traditional SKD61 material, setting a new record of 500000 consecutive injection molding without cracks.
Traditional mold steel development requires a testing cycle of over 18 months, while ForgeSteel completed the entire process from molecular structure design to forging and rolling process validation in just 6 weeks. The "sandwich" structure precision plate generated by the algorithm (surface hardness HRC60/core toughness 45J) has achieved large-scale production, especially suitable for precision stamping of titanium alloy frames in smartphones. ASML、 Foxconn and other technology manufacturers are testing the application of this material in ultra precision molds. The DeepMind team stated that the next step will be to open up API interfaces, allowing companies to customize material performance parameters, which could disrupt the global $32 billion mold steel market landscape.