当前位置:首页 > 报告详情

Optimizing Signal Integrity in Immersion-Cooled IT Platforms.pdf

上传人: 张** 编号:161434 2024-05-05 12页 1.42MB

1、Optimising Signal Integrity in immersion cooled IT platformsDr Andy E Young,CTO,AsperitasOptimising Signal Integrity in immersion cooled IT platformsPOWER AND COOLINGCOOLING ENVAs data rates increase the impact of electromagnetic properties of fluids vs.air have an impact on signal integrity.Here we

2、 will learn about what the impact is,and what we can do about it.Credit due to col-lead:Kai Wang,Technical Lead,IntelThe vast majority of servers and switches are design for use in an air-cooled environmentElectrical signals travel along transmission lines inducing electromagnetic field in surroundi

3、ng materialsThese materials cause complicationsInsertion loss(attenuation)cause reflections anddelay transmissionReview of challengeThese complications are accounted for in design(wiggles!)Bringing a multi-gigabit system to market is a major undertaking,due to stringent electrical specificationsRepl

4、acing air with dielectric fluids means potentially that good work is un-done(reduces impendence)Remaking the design for immersion is not currently and optionSo we need to knowWhat are the critical fluid properties to control,and what is the acceptable range?What methods are available to quantify the

5、se effects?What mitigation strategies are used?Review of challengeMicrostripStriplineCablesSocketCritical fluid properitesSolid SubstrateDielectric ConstantLoss Tangent,tanFR4(3GHz)40.008PTFE(3GHz)20.00028Al2O390.002Fluid ImmersionDielectric ConstantLoss Tangent,tanAir10GTL Hydrocarbon2.02 0.02Two-p

6、hase1.9 0.02OCP base spec.2.3 98%After 24 months;no changePCIe 5.0/4.0 and UPI 85%;92%;98%After 24 months and 4 monthEthernet 56G PAM4 per LanePassAfter 24 months;no changeSealed OpticalEthernetPassWIPImmersed ServersBenchmark computational approach with simple microstrip model:Measured 0.20 dBSimul

word格式文档无特别注明外均可编辑修改,预览文件经过压缩,下载原文更清晰!
三个皮匠报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
本文主要探讨了在浸没式冷却IT平台上优化信号完整性的挑战和方法。随着数据率的增加,电磁特性对信号完整性的影响日益显著。在空气冷却环境中,电气信号的传输会导致周围材料的电磁场变化,从而引起插入损耗、反射和延迟等问题。在浸没式冷却中,用介电液体替代空气,可能会影响信号的传输效率。研究显示,液体绝缘的介电常数和损耗角正切值(tanδ)对信号传输有显著影响。例如,与空气相比,某些液体如GTL羟基碳氢化合物和两相液体的介电常数和tanδ值较低,可能会对信号传输产生不同的影响。 为了量化这些影响,文中提到了使用测试和仿真方法,如通过测量设备在特定条件下的性能(插入损耗、反射损耗等),以及使用计算模型来预测信号完整性。文章还提到了优化设计策略,如限制设计中导体的暴露长度、考虑空气和液体中Socket和电缆连接器的阻抗设计、以及采用更短通道损耗设计来提高信号与噪声的比率。 最后,文章呼吁解决这些问题,并提出了测试和仿真程序,以及扩大测试设备范围和参与信号完整性社区会议等行动步骤。感兴趣的读者可以通过链接订阅更新或参与相关活动。
"液体冷却对信号完整性的影响有哪些?" "如何在液体冷却环境中优化信号完整性?" "液体冷却下的信号完整性测试与仿真方法有哪些?"
客服
商务合作
小程序
服务号
折叠