The most advanced manufacturers of computer processors are in the middle of the first big change in device architecture in a decade—the shift from finFETs to nanosheets. Another 10 years should bring about another fundamental change, where nanosheet devices are stacked atop each other to form complementary FETs (CFETs), capable of cutting the size of some circuits in half. But the latter move is likely to be a heavy lift, say experts. An in-between transistor called the forksheet might keep circuits shrinking without quite as much work.
The idea for the forksheet came from exploring the limits of the nanosheet architecture, says Julien Ryckaert, the vice president for logic technologies at Imec. The nanosheet's main feature is its horizontal stacks of silicon ribbons surrounded by its current-controlling gate. Although nanosheets only recently entered production, experts were already looking for their limits years ago. Imec was tasked with figuring out "at what point nanosheet will start tanking," he says.
Ryckaert's team found that one of the main limitations to shrinking nanosheet-based logic is keeping the separation between the two types of transistor that make up CMOS logic. The two types—NMOS and PMOS—must maintain a certain distance to limit capacitance that saps the devices' performance and power consumption. "The forksheet is a way to break that limitation," Ryckaert says.
From IEEE Spectrum
View Full Article
No entries found