...
首页> 外文期刊>Very Large Scale Integration (VLSI) Systems, IEEE Transactions on >A 10 Gbits/s/pin DFE-Less Graphics DRAM Interface With Adaptive-Bandwidth PLL for Avoiding Noise Interference and CIJ Reduction Technique
【24h】

A 10 Gbits/s/pin DFE-Less Graphics DRAM Interface With Adaptive-Bandwidth PLL for Avoiding Noise Interference and CIJ Reduction Technique

机译:具有自适应带宽PLL的10 Gbit / s / pin DFE-Less图形DRAM接口,可避免噪声干扰和CIJ降低技术

获取原文
获取原文并翻译 | 示例

摘要

A 10 Gbits/s/pin graphics DRAM interface is developed in 65-nm CMOS technology. Several design techniques are proposed for high-speed operation in a noisy environment. A fast precharging data sampler guarantees high-speed sampling without the need for a decision feedback equalizer. In order to increase the data sampling margin, the PLL bandwidth is optimized depending on the system noises, which reduces the clock jitter by up to 55.1%. The crosstalk-induced jitter (CIJ) reduction technique suppresses the DQs jitter by employing the suggested training sequence for the GDDR5 interface. Pre- and de-emphasis are merged in one auxiliary driver. This chip operates at 10 Gbits/s/pin and exhibits a data eye opening of 0.78 UI with the CIJ reduction technique. The power consumptions of the TX and RX are 8.28 and 5.5 pJ/b/channel, respectively.
机译:采用65纳米CMOS技术开发了10 Gbits / s / pin图形DRAM接口。对于在嘈杂环境中的高速运行,提出了几种设计技术。快速的预充电数据采样器可确保高速采样,而无需决策反馈均衡器。为了增加数据采样余量,根据系统噪声对PLL带宽进行了优化,从而将时钟抖动降低了55.1%。串扰引起的抖动(CIJ)降低技术通过为GDDR5接口采用建议的训练序列来抑制DQ抖动。预加重和去加重合并在一个辅助驱动器中。该芯片以10 Gbits / s / pin的速度运行,并采用CIJ减少技术展现出0.78 UI的数据眼图。 TX和RX的功耗分别为8.28和5.5 pJ / b /通道。

著录项

相似文献

  • 外文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号