摘要
为了降低静态随机存储器在处理声音和视频数据时的功耗,提出一款新型的非预充单元.相比常规6管和8管单元,其读操作消除了预充机制,抑制了无效的翻转,因此功耗得到显著优化.本单元通过多阈值技术,在保证低电压区域读噪声容限的同时也加强了数据读出的能力.而且通过引入切断反馈环的机制,有效地提升了单元写能力.此外,在存储阵列中应用半斯密特反相器,大幅地提升了静态随机存储器读操作的性能.基于SMIC 130nm工艺,分别实现容量为6kbit的非预充和常规8管静态随机存储器测试芯片.测试和仿真数据表明,这种新型存储器相比常规8管存储器在功耗的抑制上具有显著优势,可以作为低功耗应用的良好选择.
In order to save the power consumed by the static random access memory when it deals with voice or video data, a novel memory cell is proposed which eliminates the precharge mechanism in the read operation, thus suppressing the invalid power compared with the conventional 6T and 8T cell. Furthermore, multiple threshold technology is employed in the cell, which not only guarantees the read static noise margin, but also enhances read ability. In addition, a Sehmitt trigger based inverter is also applied in the cell array and hence it improves the read speed. Two memories including the proposed one and the conventional one are fabricated in 130 nm process, respectively. Test results indicate that the proposed memory is excellent in reducing power consumption compared with the conventional 8T memory, and hence it becomes a suitable choice for the low power chips.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
2018年第1期106-111,共6页
Journal of Xidian University
基金
中国科学院先导专项资助项目(XDA06020401)
国家自然科学基金资助项目(61306039)
关键词
静态随机存储器
非预充
声音和视频
低功耗
static random access memory
non-precharged
voice and video
low power