The process is optimized for artificial intelligence (AI) accelerators and processors in both training and inferencing. 12LP+ includes an updated standard cell library, an interposer for 2.5D packaging, and 0.5V SRAM bitcell that supports power-efficient movement of data between processor and memory.
The 12LP+ process builds upon Globalfoundries' established 12LP process, of which GF has shipped more than one million wafers. The 12LP process is being used by Enflame, Tenstorrent and others for AI accelerators (see Globalfoundries aids launch of Chinese AI startup ). Globalfoundries said it has developed 12LP+ to provide greater differentiation and increased value for designers in the AI space while minimizing their development and production costs.
The benefits include a 20 percent logic performance boost over 12LP and a 10 percent less area for logic.
The process has been technically qualified and is ready for production at Globalfoundries Fab 8 in Malta, New York. The company said several 12LP+ tape outs are scheduled for 2H20.
"Artificial intelligence is on a trajectory to become the most disruptive technology of our lifetime," said Amir Faintuch, senior vice president at Globalfoundries, in statement. "It is increasingly clear that the power efficiency of AI systems – in particular how many operations you can wrest from a watt of power – will be among the most critical factors a company considers when deciding to invest in data centers or edge AI applications. Our new 12LP+ solution tackles this challenge head-on. It has been engineered and optimized, obsessively so, with AI in mind."
Related links and articles: