Abstract

In 2009, we discovered an oxide semiconductor (OS) of a ceramic material that has a novel crystal morphology, a c-axis aligned crystalline (CAAC) structure, which is different from a single crystal structure, a polycrystalline structure, and an amorphous structure. Since then, we have evaluated the compatibility of LSIs using OS (OSLSIs) with LSIs using Si (SiLSIs) in terms of manufacturing apparatuses and lines. Consequently, we found that OSLSIs excel SiLSIs in some characteristics, and that the stable production of OSLSIs requires controlling the following three key points: concentrations of hydrogen and oxygen, and distortion energy. Furthermore, we succeeded in downscaling OSLSIs, which is unfeasible for SiLSIs. We obtained a cutoff frequency of 120 GHz in an OS field-effect transistor (OSFET) having a gate length of 6.7 nm. We also obtained an Ioff of 10−22 A/μm (R T.) in an OSFET having a gate length of 22 nm, which is lower than that of a Si field-effect transistor by ten orders of magnitude. Additionally, a stacked structure in which an OSLSI and a SiLSI are combined enabled a normally-off central processing unit (NoffCPU) to have an ultralow power consumption of 10 TOPS/W (8 bits). These characteristics attract some companies that positively consider the mass production of the OSLSIs. The advancement in OSLSI technology can realize the application to data centers and supercomputers. We believe that an ultralow power LSI that employs an OS/Si combined (3D) structure and has a power consumption of one-hundredth of a conventional LSI is the one and only solution to global warming in the topic discussed. Therefore, we propose an OS as a next-generation VLSI semiconductor material.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call