In dieser Arbeit werden Quantenpunkt-Speichertransistoren basierend auf modulationsdotierten GaAs/AlGaAs Heterostrukturen mit vorpositionierten InAs Quantenpunkten vorgestellt, welche in Abhängigkeit der Ladung auf den Quantenpunkten unterschiedliche Widerstände und Kapazitäten aufweisen. Diese Ladungsabhängigkeiten führen beim Anlegen von periodischen Spannungen zu charakteristischen, durch den Ursprung gehenden Hysteresen in der Strom-Spannungs- und der Ladungs-Spannungs-Kennlinie. Die ladungsabhängigen Widerstände und Kapazitäten ermöglichen die Realisierung von neuromorphen Operationen durch Nachahmung von synaptischen Funktionalitäten und arithmetischen Operationen durch Integration von Spannungs- und Lichtpulsen. / In this thesis, state-dependent resistances and capacitances in quantum dot floating gate transistors based on modulation doped GaAs/AlGaAs heterostructures with site-controlled InAs quantum dots are presented. The accumulation of electrons in the quantum dots simultaneously increases the resistance and decreases the capacitance, which leads to characteristic pinched hysteresis loops in the current-voltage- and the charge-voltage-characteristics when applying periodic input signals. The concurrent resistance and capacitance switching enables the realization of neuromorphic operations via mimicking of synaptic functionalities and arithmetic operations via the integration of voltage and light pulses.

Nanokristalle werden beispielsweise für eine Anwendung in Solarzellen, Lichtemittern und nichtflüchtigen Datenspeichern diskutiert. Damit diese Anwendungen funktionieren können, ist eine genaue Kontrolle der Kristallitgröße sowie der Flächendichte und Lage der Kristallite in der Matrix wichtig. Zudem sollte die Matrix amorph sein, da amorphe Matrixmaterialien die Nanokristall-Oberfläche besser passivieren und beständiger gegen Leckströme sind. In dieser Arbeit werden Ge-Nanokristalle in die Hoch-Epsilon-Dielektrika ZrO2 und TaZrOx eingebettet. Im System Ge/ZrO2 kristallisieren die Ge-Cluster und die ZrO2-Matrix bei der gleichen Temperatur. Aufgrund der kristallinen Matrix weicht die Form der Ge-Nanokristalle von einer Kugel ab, worunter unter anderem die Größenkontrolle leidet. Die Beimischung von Ta2O5 stabilisiert die amorphe Phase des ZrO2 und verhindert dadurch die gemeinsame Kristallisation. Dadurch wird es im System Ge/TaZrOx möglich, kugelförmige Ge-Nanokristalle im Größenbereich von 3 nm bis 6 nm positionskontrolliert in eine amorphe Matrix einzubetten. Für die Untersuchung einer möglichen Anwendung des Materialsystems wurden Speicherzellen eines nichtflüchtigen Datenspeichers auf Basis von Ge-Nanokristallen hergestellt. Dabei zeigte sich, dass das System Ge/TaZrOx überdurchschnittlich viele Ladungen speichert und daher für diese Anwendung vielversprechend ist. Zudem stabilisiert die Beimischung von Ta2O5 eine extrem seltene orthorhombische Modifikation des ZrO2. Für ferroelektrische Datenspeicher könnte diese Phase eine aussichtsreiche Alternative zum HfO2 sein.

Database systems have long been optimized to hide the higher latency of storage media, yielding complex persistence mechanisms. With the advent of large DRAM capacities, it became possible to keep a full copy of the data in DRAM. Systems that leverage this possibility, such as main-memory databases, keep two copies of the data in two different formats: one in main memory and the other one in storage. The two copies are kept synchronized using snapshotting and logging. This main-memory-centric architecture yields nearly two orders of magnitude faster analytical processing than traditional, disk-centric ones. The rise of Big Data emphasized the importance of such systems with an ever-increasing need for more main memory. However, DRAM is hitting its scalability limits: It is intrinsically hard to further increase its density. Storage-Class Memory (SCM) is a group of novel memory technologies that promise to alleviate DRAM’s scalability limits. They combine the non-volatility, density, and economic characteristics of storage media with the byte-addressability and a latency close to that of DRAM. Therefore, SCM can serve as persistent main memory, thereby bridging the gap between main memory and storage. In this dissertation, we explore the impact of SCM as persistent main memory on database systems. Assuming a hybrid SCM-DRAM hardware architecture, we propose a novel software architecture for database systems that places primary data in SCM and directly operates on it, eliminating the need for explicit IO. This architecture yields many benefits: First, it obviates the need to reload data from storage to main memory during recovery, as data is discovered and accessed directly in SCM. Second, it allows replacing the traditional logging infrastructure by fine-grained, cheap micro-logging at data-structure level. Third, secondary data can be stored in DRAM and reconstructed during recovery. Fourth, system runtime information can be stored in SCM to improve recovery time. Finally, the system may retain and continue in-flight transactions in case of system failures. However, SCM is no panacea as it raises unprecedented programming challenges. Given its byte-addressability and low latency, processors can access, read, modify, and persist data in SCM using load/store instructions at a CPU cache line granularity. The path from CPU registers to SCM is long and mostly volatile, including store buffers and CPU caches, leaving the programmer with little control over when data is persisted. Therefore, there is a need to enforce the order and durability of SCM writes using persistence primitives, such as cache line flushing instructions. This in turn creates new failure scenarios, such as missing or misplaced persistence primitives. We devise several building blocks to overcome these challenges. First, we identify the programming challenges of SCM and present a sound programming model that solves them. Then, we tackle memory management, as the first required building block to build a database system, by designing a highly scalable SCM allocator, named PAllocator, that fulfills the versatile needs of database systems. Thereafter, we propose the FPTree, a highly scalable hybrid SCM-DRAM persistent B+-Tree that bridges the gap between the performance of transient and persistent B+-Trees. Using these building blocks, we realize our envisioned database architecture in SOFORT, a hybrid SCM-DRAM columnar transactional engine. We propose an SCM-optimized MVCC scheme that eliminates write-ahead logging from the critical path of transactions. Since SCM -resident data is near-instantly available upon recovery, the new recovery bottleneck is rebuilding DRAM-based data. To alleviate this bottleneck, we propose a novel recovery technique that achieves nearly instant responsiveness of the database by accepting queries right after recovering SCM -based data, while rebuilding DRAM -based data in the background. Additionally, SCM brings new failure scenarios that existing testing tools cannot detect. Hence, we propose an online testing framework that is able to automatically simulate power failures and detect missing or misplaced persistence primitives. Finally, our proposed building blocks can serve to build more complex systems, paving the way for future database systems on SCM.

Logic-in-memory circuits promise to overcome the von-Neumann bottleneck, which constitutes one of the limiting factors to data throughput and power consumption of electronic devices. In the following we present four-input logic gates based on only two ferroelectric FETs (FeFETs) with hafnium oxide as the ferroelectric material. By utilizing two complementary inputs, a XOR and a XNOR gate are created. The use of only two FeFETs results in a compact and nonvolatile design. This realization, moreover, directly couples the memory and logic function of the FeFET. The feasibility of the proposed structures is revealed by electrical measurements of HKMG FeFET memory arrays manufactured in 28nm technology.

During the last decade ferroelectrics based on doped hafnium oxide emerged as promising candidates for realization of ultra-low-power non-volatile memories. Two spontaneous polarization states occurring in the material that can be altered by applying electrical fields rather than forcing a current through and the materials compatibility to CMOS processing are the main benefits setting the concept apart from other emerging memories. 1T1C ferroelectric random access memories (FeRAM) as well as 1T FeFET concepts are under investigation. In this article the application of hafnium based ferroelectric memories in different flavours and their ranking in the memory landscape are discussed.

Database systems have long been optimized to hide the higher latency of storage media, yielding complex persistence mechanisms. With the advent of large DRAM capacities, it became possible to keep a full copy of the data in DRAM. Systems that leverage this possibility, such as main-memory databases, keep two copies of the data in two different formats: one in main memory and the other one in storage. The two copies are kept synchronized using snapshotting and logging. This main-memory-centric architecture yields nearly two orders of magnitude faster analytical processing than traditional, disk-centric ones. The rise of Big Data emphasized the importance of such systems with an ever-increasing need for more main memory. However, DRAM is hitting its scalability limits: It is intrinsically hard to further increase its density. Storage-Class Memory (SCM) is a group of novel memory technologies that promise to alleviate DRAM’s scalability limits. They combine the non-volatility, density, and economic characteristics of storage media with the byte-addressability and a latency close to that of DRAM. Therefore, SCM can serve as persistent main memory, thereby bridging the gap between main memory and storage. In this dissertation, we explore the impact of SCM as persistent main memory on database systems. Assuming a hybrid SCM-DRAM hardware architecture, we propose a novel software architecture for database systems that places primary data in SCM and directly operates on it, eliminating the need for explicit IO. This architecture yields many benefits: First, it obviates the need to reload data from storage to main memory during recovery, as data is discovered and accessed directly in SCM. Second, it allows replacing the traditional logging infrastructure by fine-grained, cheap micro-logging at data-structure level. Third, secondary data can be stored in DRAM and reconstructed during recovery. Fourth, system runtime information can be stored in SCM to improve recovery time. Finally, the system may retain and continue in-flight transactions in case of system failures. However, SCM is no panacea as it raises unprecedented programming challenges. Given its byte-addressability and low latency, processors can access, read, modify, and persist data in SCM using load/store instructions at a CPU cache line granularity. The path from CPU registers to SCM is long and mostly volatile, including store buffers and CPU caches, leaving the programmer with little control over when data is persisted. Therefore, there is a need to enforce the order and durability of SCM writes using persistence primitives, such as cache line flushing instructions. This in turn creates new failure scenarios, such as missing or misplaced persistence primitives. We devise several building blocks to overcome these challenges. First, we identify the programming challenges of SCM and present a sound programming model that solves them. Then, we tackle memory management, as the first required building block to build a database system, by designing a highly scalable SCM allocator, named PAllocator, that fulfills the versatile needs of database systems. Thereafter, we propose the FPTree, a highly scalable hybrid SCM-DRAM persistent B+-Tree that bridges the gap between the performance of transient and persistent B+-Trees. Using these building blocks, we realize our envisioned database architecture in SOFORT, a hybrid SCM-DRAM columnar transactional engine. We propose an SCM-optimized MVCC scheme that eliminates write-ahead logging from the critical path of transactions. Since SCM -resident data is near-instantly available upon recovery, the new recovery bottleneck is rebuilding DRAM-based data. To alleviate this bottleneck, we propose a novel recovery technique that achieves nearly instant responsiveness of the database by accepting queries right after recovering SCM -based data, while rebuilding DRAM -based data in the background. Additionally, SCM brings new failure scenarios that existing testing tools cannot detect. Hence, we propose an online testing framework that is able to automatically simulate power failures and detect missing or misplaced persistence primitives. Finally, our proposed building blocks can serve to build more complex systems, paving the way for future database systems on SCM.

With the rise of computationally expensive application domains such as machine learning, genomics, and fluids simulation, the quest for performance and energy-efficient computing has gained unprecedented momentum. The significant increase in computing and memory devices in modern systems has resulted in an unsustainable surge in energy consumption, a substantial portion of which is attributed to the memory system. The scaling of conventional memory technologies and their suitability for the next-generation system is also questionable. This has led to the emergence and rise of nonvolatile memory ( NVM ) technologies. Today, in different development stages, several NVM technologies are competing for their rapid access to the market. Racetrack memory ( RTM ) is one such nonvolatile memory technology that promises SRAM -comparable latency, reduced energy consumption, and unprecedented density compared to other technologies. However, racetrack memory ( RTM ) is sequential in nature, i.e., data in an RTM cell needs to be shifted to an access port before it can be accessed. These shift operations incur performance and energy penalties. An ideal RTM , requiring at most one shift per access, can easily outperform SRAM . However, in the worst-cast shifting scenario, RTM can be an order of magnitude slower than SRAM . This thesis presents an overview of the RTM device physics, its evolution, strengths and challenges, and its application in the memory subsystem. We develop tools that allow the programmability and modeling of RTM -based systems. For shifts minimization, we propose a set of techniques including optimal, near-optimal, and evolutionary algorithms for efficient scalar and instruction placement in RTMs . For array accesses, we explore schedule and layout transformations that eliminate the longer overhead shifts in RTMs . We present an automatic compilation framework that analyzes static control flow programs and transforms the loop traversal order and memory layout to maximize accesses to consecutive RTM locations and minimize shifts. We develop a simulation framework called RTSim that models various RTM parameters and enables accurate architectural level simulation. Finally, to demonstrate the RTM potential in non-Von-Neumann in-memory computing paradigms, we exploit its device attributes to implement logic and arithmetic operations. As a concrete use-case, we implement an entire hyperdimensional computing framework in RTM to accelerate the language recognition problem. Our evaluation shows considerable performance and energy improvements compared to conventional Von-Neumann models and state-of-the-art accelerators.

Ferroelectric field effect transistor (FeFET) memories based on a new type of ferroelectric material (silicon doped hafnium oxide) were studied within the scope of the present work. Utilisation of silicon doped hafnium oxide (Si:HfO2) thin films instead of conventional perovskite ferroelectrics as a functional layer in FeFETs provides compatibility to the CMOS process as well as improved device scalability. The influence of different process parameters on the properties of Si:HfO2 thin films was analysed in order to gain better insight into the occurrence of ferroelectricity in this system. A subsequent examination of the potential of this material as well as its possible limitations with the respect to the application in non-volatile memories followed. The Si:HfO2-based ferroelectric transistors that were fully integrated into the state-of-the-art high-k metal gate CMOS technology were studied in this work for the first time. The memory performance of these devices scaled down to 28 nm gate length was investigated. Special attention was paid to the charge trapping phenomenon shown to significantly affect the device behaviour.

Im Rahmen dieser Arbeit werden haftstellen-basierte Speicherzellen als mögliche Alternative zum bestehenden Floating-Gate Konzept untersucht. Hierbei wird zunächst mittels Simulation und ausgewählten Messverfahren das Verständnis der Funktionsweise vertieft. Der darauffolgende Abschnitt befasst sich mit der Verbesserung der elektrischen Eigenschaften, basierend auf Änderungen der verwendeten Materialien und dem räumlichen Aufbau. Abschließend erfolgt die Untersuchung der Anwendbarkeit des Zellkonzeptes in hochdichten Zellenfeldern.

Im Rahmen dieser Arbeit werden haftstellen-basierte Speicherzellen als mögliche Alternative zum bestehenden Floating-Gate Konzept untersucht. Hierbei wird zunächst mittels Simulation und ausgewählten Messverfahren das Verständnis der Funktionsweise vertieft. Der darauffolgende Abschnitt befasst sich mit der Verbesserung der elektrischen Eigenschaften, basierend auf Änderungen der verwendeten Materialien und dem räumlichen Aufbau. Abschließend erfolgt die Untersuchung der Anwendbarkeit des Zellkonzeptes in hochdichten Zellenfeldern.:Kurzfassung Abstract 1 Einleitung 2 Grundlagen aktiver Halbleiterelemente 2.1 Die MOS-Struktur 2.2 Der MOS-Feldeffekt-Transistor 2.3 Nichtflüchtige Festkörperspeicher 2.4 Speicherarchitekturen 2.5 Charakterisierungsmethoden von Halbleiter-Speicherelementen 3 Defektbasierte Ladungsspeicherung in dielektrischen Schichten 3.1 Physikalische Grundlagen von Haftstellen 3.2 Betrachtung der vertikalen Ladungsverteilung mit Hilfe von Simulationen 3.3 Ableitung der vertikalen Ladungsverteilung aus Messungen 4 Elektrisches Verhalten einer haftstellen-basierten Speicherzelle 4.1 Auswirkung von inhomogen verteilter Ladung in der Speicherschicht 4.2 Auswirkungen von Al2O3-Topoxid auf das Zellverhalten 4.3 Auswirkung des Steuerelektrodenmaterials auf das Zellverhalten 4.4 Einfluss von Kanal- und Source/Drain-Dotierung 5 Integration in eine stark skalierte NAND Architektur 5.1 Auswirkung struktureller Effekte auf die Speicherzelle 5.2 Störmechanismen beim Betrieb von stark skalierten NAND-Speichern 6 Zusammenfassung und Ausblick 6.1 Zusammenfassung 6.2 Ausblick Danksagung Lebenslauf Symbol- und Abkürzungsverzeichnis Literaturverzeichnis