Wafer-scale integration
Using a whole wafer as one giant chip instead of cutting it into dies, as Cerebras does for its AI processors.
Also: wafer-scale engine, WSE
Since the 1980s engineers have tried to build one chip from a whole wafer, which avoids the slow links between separate chips. Early efforts such as Trilogy Systems failed on yield, because a wafer always contains defects. Cerebras Systems shipped its first Wafer Scale Engine in 2019, about 46,000 mm² and 1.2 trillion transistors, using spare cores and routing around faults, and has followed with later generations. Tesla's Dojo and TSMC's system-on-wafer packaging take a related approach by joining many dies across a wafer-sized carrier.