It was the end of the semiconductor roadmap for 193-nanometer lithography--the next generation would have to go to 157-nanometer light.
Jim Blatchford, Texas Instrument's manager of front-end processing, had just finished negotiating for a leading edge 157-nanometer photolithography system, but he was still worried about the unproven technology, so he went to an SPIE (Society of Photo-Optical Instrumentation Engineers) session on 157-nanometer lithography.
When he arrived, however, he was surprised to find the room nearly empty. The rain was pouring down outside making a lonley pitter patter without the usual crowd noises to mask it. Something was wrong at 157-nanometer.
"I remember it so well. We had just finished negotiations on our 157-nanometer tool, but when I walked into a 157-nanmoter session at SPIE 2004 the room was almost completely empty. The sound of the rain pouring down on the roof was eery. I knew right away that something was wrong," said Blatchford.
The mystery was solved when Blatchford found out that the attendees were crowded into the immersion lithography session for 193-nanometer, where presenters were claiming that engineers didn't need to risk going to 157-nanometer. All they had to do was immerse the reduction lens nearest the wafer in water to decrease feature sizes by its index-of-refractionn (1.44). And with higher index fluids, it should be possible to extend 193-nanometer lithography evern further.
"What was most surprising to me was just how quickly everyone adopted immersion lithography. Once it was established as a feasible technology, everybody got on-board immediately," said Blatchford.
Click on image to enlarge.
Light at 193 nm has a resolution limit on the order of 65 nm, but
immersion lithography uses the same light source and yet realizes a
semiconductor process at precisions as high as 40-45nm. (Source: Nikon)
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.