Logo image
相容於邏輯製程之淺溝槽絕緣邊界電阻式記憶體
Thesis

相容於邏輯製程之淺溝槽絕緣邊界電阻式記憶體

廖矩鋒
Masters, 國立清華大學, 電子工程研究所
2015

Abstract

電阻式記憶體 淺溝槽絕緣 RRAM Shallow Trench Isolation
In the future, digitized world can be achieved by IOT. All item will be network links, human life may be changed. With the wisdom handheld products takeoff, products are slim and light as a trend and Process-in-Memory, PIM is the most representative technology to make the memory embedded in CPU. On the one hand it can enhance the operation speed, on the other hand it can reduce the power consumption. This technology might be a key to the generation of IOT and big data. Large information must have a corresponding storage media for operation and usage. Currently, the main NVM is flash on the market. As semiconductor manufacturing process scales, the charge in the floating gate decreases .The storage state will become difficult to identify. Flash may encounter the physical limitation that can’t operate and use normally. Therefore, new NVM exploitation is important nowadays. Based on 40nm COMS logic process, we realized Shallow trench Isolation Sidewall Edge Resistive Random Access Memory(STI Sidewall Edge RRAM).No additional process masks and ultra-small size is the characteristic of this memory. We can control the RRAM film easily and precisely. The variation of this cell may decrease with process scaling. By DC and AC operation, this memory has the feature of low power consumption and high program speed. Incremental Step Pulse Programming(ISPP) Algorithm can be used to promote the endurance to make this memory achieve 1000K Set/Reset cycles. Data stored show stability under 150oC baking over a period of 1000K seconds and it can endure 10K seconds continuously reading that guarantee STI Sidewall Edge RRAM unit has no reliability concern. STI Sidewall Edge RRAM has great operation feature and data retention ability.

Metrics

1 Record Views

Details

Logo image