当前位置: X-MOL 学术IEEE Comput. Archit. Lett. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
DAM: Deadblock Aware Migration Techniques for STT-RAM-Based Hybrid Caches
IEEE Computer Architecture Letters ( IF 1.4 ) Pub Date : 2021-04-09 , DOI: 10.1109/lca.2021.3071717
Arindam Sarkar , Newton Singh , Varun Venkitaraman , Virendra Singh

Transverse mode instability (TMI) limits the power scaling of high-power fiber lasers and amplifiers. One method of mitigating TMI is to strip unwanted modes by bending the fiber. However, the bending-induced mode distortion will change the TMI coefficient and influences the TMI threshold. In this work, we analyzed the bending effects on the mode profiles by inducing different refractive index tilts corresponding to different bending radii. With these mode profiles and with different bending conditions, a modified TMI coefficient was calculated to analyze the bending effects on TMI for different fibers. The simulation results show that tighter bending can lead to a somewhat lower TMI coefficient and slightly improve the TMI threshold.

中文翻译:


DAM:基于 STT-RAM 的混合缓存的死块感知迁移技术



横模不稳定性 (TMI) 限制了高功率光纤激光器和放大器的功率缩放。缓解 TMI 的一种方法是通过弯曲光纤去除不需要的模式。然而,弯曲引起的模式畸变会改变TMI系数并影响TMI阈值。在这项工作中,我们通过引入与不同弯曲半径相对应的不同折射率倾斜来分析弯曲对模式轮廓的影响。利用这些模式分布和不同的弯曲条件,计算修正的 TMI 系数,以分析不同光纤的弯曲对 TMI 的影响。仿真结果表明,更紧的弯曲可以导致 TMI 系数稍低,并略微提高 TMI 阈值。
更新日期:2021-04-09
down
wechat
bug