Daniele Ielmini, Ph.D.
相变存储器元件及阵列之可靠性议题及建模
经过工业领域和学术领域近十五年的研究和发展，不论是独立型或是嵌入式的存储器，相变存储器（PCM）已成为最成熟的新兴技术。对PCM商品化而言，不可或缺的是制程与可靠性的严格控制。本次报告着眼于PCM可靠性的影响机制和变动模式，探讨在单一存储元件以及存储器阵列层次的电阻漂移和结晶情况。
发生在复位操作之后PCM电阻持续增加的情形，被称为电阻漂移。电阻漂移起因于电阻处在非晶相结构时的结构变松散。非晶相结构缺陷经历加温会加速产生，如针对此现象建立精确的模型，就能够描述电阻和临界电压和的漂移与温度等等。本次演讲将回顾漂移的物理模型，并探讨读取储存器单元的方式来减少电阻漂移对多阶存储单元的影响。
After about 15 years of research and development in the industry and academia, phase change memory (PCM) has become the most mature emerging technology for both stand-alone and embedded memory. For PCM commercialization, however, a tight control of process and reliability is essential. This talk will address reliability mechanisms and models in PCM, discussing resistance drift and crystallization at both single device and array levels.
Resistance drift, namely the increase of resistance after reset operation, is due to the structural relaxation of the amorphous phase. By accurate modeling of the T-accelerated annealing of defects in the amorphous structure, it is possible to describe drift of both resistance and threshold voltage as a function of time. This talk will review physical models of drift and discuss read schemes to minimize the impact of resistance drift in multilevel cell.
The amorphous state of PCM is also prone to crystallization, resulting in a drop of resistance above the crystallization temperature Tx. This talk will summarize Arrhenius models to predict the Tdependent data retention in both single PCM devices and multiGb arrays. To enhance PCM retention at high temperature, e.g., for automotive and embedded applications, the phase change material must be optimized to achieve high Tx. However, high-Tx materials generally display drift of the set (crystalline) state, which might also contribute to retention failure. This talk will review reliability characteristics and modeling of Ge-rich Ge2Sb2Te5, highlighting the tradeoff between reset-state and set-state retention in embedded PCM.
Chung H. Lam, Ph.D.
神经形态工程学
当 CMOS 硅晶持续微缩到电路元件的一个临界尺寸，并到达一个只有个位数纳米尺寸时，利用范纽曼架构设计的微处理器设计可以达到功率及效能限制。在早于2000年，没有增加核心操作频率的多核心处理器和平行处理器，己经被用来延伸功率及效能限制，以保持摩尔定律的可行性。然而，阿姆达尔定律 (Amdahl’s Law) 争论利用平行处理器来提升效能是取决于需要被依序处理演算法的比例，进化已经提供给我们拥有最有效的平行处理架构 ——生物脑。本次演讲中，我们将会检验什么是我们可以做的，利用我们所知的一些大脑如何运作原理去设计机器来模仿大脑的运作。
Microprocessors designed with von Neumann architecture are hitting the power and performance limits as silicon CMOS continues to scale the critical dimensions of the circuit components towards single digit nanometer size limit. Multi-core processor, parallel processing without increasing operating frequency of the cores, was introduced in the early 2000 to extend the power and performance scaling, keeping Moore’s Law viable. Amdahl’s Law, however, argues that the performance speedup with parallel processing is governed by the percentage of algorithm that needed be serial. Evolution has provided us with the most efficient parallel processing architecture: the biological brain. In this talk, we shall examine what we can do with little that we know about how the brain works to design machines to mimick the brain.
Ming Li, Ph.D.
后摩尔时代的硅纳米线晶体管
在本次研讨会中，将探讨近几年来硅纳米线晶体管技术在北京大学的研发进展，并详细讨论硅纳米线晶体管(SNWT)的制造技术及其性能上所遇到的挑战。可极端缩放的硅纳米线晶体管(SNWT)已证实在短通道效应和高性能要求之下所呈现的超级可控性。此外也将讨论迁移率的退化、寄生电阻、变异性和可靠性方面的挑战以及硅纳米线晶体管(SNWT)在逻辑和存储器技术的未来潜力。
Silicon nanowire transistor technologies in recent years and the progress in Peking University will be reviewed in this talk. The fabrication technologies and performance challenges for SNWT are discussed in details. The super controllability on short-channel effect and high performance are demonstrated by an extremely scaled SNWT. The challenges such as mobility degradation, parasitic resistance, variability and reliability will also be discussed as well the future potential of SNWT in logic and memory technology.
Meng-Fan Chang, Ph.D.
电阻式存储器电路和应用的挑战
存储器已成为目前穿戴式装置、云端伺服器以及大数据处理器达到低能量损耗的瓶颈之一，电阻式存储器是目前达到高能量效率的非挥发性存储器的条件，本次演讲将讨论电阻式存储器在电路设计及应用上的挑战和趋势，并以一些晶片设计为实例呈现；包含高速度、高效益、及低电压电阻式存储器为一体设计。另外，在新的应用方面，例如非挥发性静态随机存取存储器器 (nvSRAM)，非挥发性逻辑(nonvolatile logics)，以及非挥发性三态内容定址存储器(nonvolatile TCAM)，在无线传感器网络及大数据处理器的应用也会在此次的演讲中阐述。
Memory has become one of the bottlenecks for wearable devices, cloud servers, and big-data processing to achieve low energy consumption. Resistive-type memories are the candidates for energy-efficient nonvolatile memory. This talk will discuss the trends and design challenges in circuits and applications for resistive-type memories. Several silicon examples will be presented, including high-speed, area-efficient, and low-voltage resistive-type memory macros. New applications, such as nonvolatile-SRAM (nvSRAM), nonvolatile logics, and nonvolatile TCAM
(nvTCAM) in wireless sensor network and big-data processors will also be presented in this talk.
Koukou Suu, Ph.D.
ULVAC于相变存储器与尖端非挥发存储器的制造技术开发
相变存储器(PCRAM) 和其他尖端非挥发存储器 (NVMs)在存储类内存(SCM)和智能类内存正引起高度关注，主要是其低功率电压和快速存取功能。在尖端非挥发存储器(NVM)制造最大挑战是新材料的处理，如多元素相变化材料、氧化物与制程堆栈，都是非挥发存储器(NVM)必须材料与技术。但目前这些材料与技术并不存于现有集成电路(LSI) 制程中。
ULVAC一直致力于相变存储器(PCRAM) 和其他尖端非挥发存储器 (NVMs)的开发制造解决方案，自2004年来已在客户工厂安装众多量产与研究设备，这些设备主要用于非挥发存储器制造。这次演讲中，我们将针对相变存储器(PCRAM) 和其他尖端非挥发存储器 (NVMs)最新开发状态进行报告。
Phase Change Memory (PCRAM) and other emerging non-volatile memories (NVMs) are catching stronger and stronger attention for their applications as Storage Class Memory (SCM) and Smart Memory due to their low-power/voltage and fast operation features. One of the biggest challenges in manufacturing NVMs is how to process the new materials such as multi-elemental phase change materials, oxides and their stacks which are necessary to realize the functionalities of the NVMs but are not existing materials in current LSI.
ULVAC has been developing manufacturing solutions including both manufacturing tools and processes for PCRAM and other NVMs since 2004 and successfully installed numerous tools to our customers for mass-production and R&D of NVMs . In this talk, we would like to report our recent development status for PCRAM and other NVMs.
Luca Perniola, Ph.D.
颠覆性技术在嵌入式存储器的应用
后段制程式存储器的颠覆技术已经可以有效取代现今主流的快闪式存储器。在过去学术发表里对于存储器的资料保持力以及在回流焊接上的问题也有着优秀的学术成果。这次发表将会着重于由130纳米到28纳米，相变存储器、氧化物电阻式存储器和导电桥接随机存取存储器的技术成就。
Disruptive technologies like BEOL memories are becoming valid alternative to the current mainstream consisting of eFlash. In the last years outstanding advances have been claimed in the literature especially focusing on the advances on data retention and in particular the soldering reflow issue. In this presentation the last achievements on optimized memory stacks consisting of PCRAM, OxRAM and CBRAM will be presented from 130 nm down to 28 nm node.
Geoffrey W. Burr, Ph.D.
储存级存储器及非Von Neumann运算的交叉阵列
50多年来，Von Neumann 信息处理系统–即一种藉由"存储“传达指令并经由指定的中央处理单元执行指令的能力已被大幅的提升。这个可圈可点的历史，过去一直是被一种永远不断增加密度的摩尔定律(Moore’s Law) 所驱动，实际驱动者为Dennard’s 定律(Dennard’s Law)：一种元件微缩方法，使得每一个更小的电晶体的效能更胜于前一代。不过，Dennard’s 定律几年以前已经失效，Moore’s 定律 (Moore’s Law) 近年来已趋于缓慢。为了找寻持续提升运算系统的效能，信息(IT)产业的焦点已转移到非Von Neumann演算法，由人脑所衍生的运算架构。
同时，存储器技术也进入快速变革的年代，新一代的非挥发存储器 (NVM) ；如相变存储器(PCM)、电阻式存储器(RRAM)、及自旋力矩磁性存储器(STT-RAM)，熔合了传统三种存储器技术；静态随机存储器 (SRAM)、动态随机存储器(DRAM)、及快闪存储器(FLASH)的优缺点。此新式存储器技术将可达成我们所追求的储存级存储器 (SCM)，结合如固态存储器的高效能及如磁性硬碟片的耐久性，低成本特性。
此非挥发性存储器阵列藉由其可变化的电阻特性相似于脑神经突出元，亦可应用于非Von Neumann 脑神经运算系统。这是非常吸引人的应用，较多的神经突出元将有助于处理运算的良率及其变异容忍范围。这个领域仍然需要投入高质量的研究工作，以突破应用现况。
此次将讨论我们把大型非挥发存储器交叉阵列在这些应用上的研究成果。简单的探讨早期的相变存储器(PCM)、储存级存储器(SCM)、及含铜的混合式离电导电(MIEC)元件后，再探讨目前所关注的以非挥发存储器为基础的脑神经系统工程议题。
For more than 50 years, the capabilities of Von Neumann-style information processing systems — in which a "memory" delivers operations and then operands to a dedicated "central processing unit" — have improved dramatically. While it may seem that this remarkable history was driven by everincreasing density (Moore's Law), the actual driver was Dennard's Law: a device-scaling methodology which allowed each generation of smaller transistors to actually perform better, in every way, than the previous generation. Unfortunately, Dennard's Law terminated some years ago, and as a result, Moore's Law is now slowing considerably. In a search for ways to continue to improve computing systems, the attention of the IT industry has turned to Non-Von Neumann algorithms, and in particular, to computing architectures motivated by the human brain.
At the same time, memory technology has been going through a period of rapid change, as new nonvolatile memories (NVM) — such as Phase Change Memory (PCM), Resistance RAM (RRAM), and Spin-Torque-Transfer Magnetic RAM (STT-MRAM) — emerge that complement and augment the traditional triad of SRAM, DRAM, and Flash. Such memories could enable Storage-Class Memory (SCM) — an emerging memory category that seeks to combine the high performance and robustness of solid-state memory with the long-term retention and low cost of conventional harddisk magnetic storage.
Such large arrays of NVM can also be used in non-Von Neumann neuromorphic computational schemes, with device conductance serving as the plastic (modifiable) “weight” of each “native” synaptic device. This is an attractive application for these devices, because while many synaptic weights are required, requirements on yield and variability can be more relaxed. However, work in this field has remained highly qualitative in nature, and slow to scale in size.
I will discuss our recent work towards large crossbar arrays of NVM for both of these applications.
After briefly reviewing earlier work on PCM, SCM, and access devices based on copper-containing Mixed-Ionic-Electronic-Conduction (MIEC), I will discuss our recent work on quantitatively assessing the engineering tradeoffs inherent in NVM-based neuromorphic systems.
Ming Hsieh, Member of NAE
时代全芯 – 新一代存储技术制造先锋
时代全芯公司正在宁波兴建一座12吋的PCM晶圆厂。演讲中将阐述公司成立的愿景，同时将时代全芯公司目前晶圆厂兴建的阶段、公司里程的记录及未来策略的规划进行报告。
Advanced Memory Technology (AMT) Corp is constructing a 12” fab in Ningbo for PCM productization. In this talk, we will briefly describe the founding of the company, its organizational structure. We will also present the fab start-up schedule, major operational milestones and business strategies.