摘要
为了加快微处理器中线性地址向物理地址转换的速度,提出了一种高速TLB结构。结构采用全定制的CAM阵列和SRAM阵列,并根据CAM和SRAM单元的输出特点设计了精巧的读出放大逻辑,有效提高了TLB的读出速度。经流片测试,表明设计正确可靠,能够保证地址转换延时在1ns左右。
A new high-speed TLB architecture is designed for accelerating the address transition rate from linear address to physical one in micro processors.Full custom circuit parts of CAM and SRAM are adopted. According to the output signal characteristic of the tow kinds of storage units ,amplifying and reading circuits are elaborately designed to improve the reading speed of TLB. Taped out chips can work correctly and reliably,and can keep the transition delay at about 1 ns.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第16期1-3,86,共4页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of China under GrantNo.60473079)