Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 20 for “"eDRAM"”.
-
Compute in eDRAM using indium gallium zinc oxide transistors
… embedded Dynamic Random Access Memory (eDRAM) making use of backend of the line (BEOL) integrated C-Axis Aligned Crystalline (CAAC) Indium Gallium Zinc Oxide (IGZO) transistors can enable energy efficient CIM compute. The extreme low leakage of these IGZO transistors can be used to …
-
Process variation and scaling effects on embedded DRAMs.
… trend in the power comparison between SRAM and eDRAM is conducted. To improve the performance of eDRAM various kinds of eDRAM cells are introduced. The scaling trend of major leakage currents affecting eDRAM performance is presented, as well as their impact on the behavior of various cells. A …
-
The Development of Embedded DRAM Statistical Quality Models at Test and Use Conditions
… use of embedded Dynamic Random Access Memory (eDRAM) is increasing in our electronics that require large memories, such as gaming consoles and computer network routers. Unlike external DRAMs, eDRAMs are embedded inside ASICs for faster read and write operations. Until recently, eDRAMs required …
-
Refresh reduction in dynamic memories
… a low-leakage technology such as embedded DRAM (eDRAM). Unfortunately, eDRAM, being a dynamic memory, requires periodic refresh, which ends up consuming substantial energy for large last-level caches. In upcoming architectures that stack a processor die and multiple DRAM dies, DRAM dies …
-
Technology Implications for Large Last-Level Caches
… SRAM, STT-RAM (MRAM), and embedded DRAM (eDRAM), have been used or considered as the technology to implement L3Cs. However, each of them has inherent weaknesses: SRAM is relatively low density and dissipates high leakage; STT-RAM has long write latency and requires high write energy; eDRAM …
-
Refrint: intelligent refresh to minimize power in on-chip multiprocessor cache hierarchies
… a low-leakage technology such as embedded DRAM (eDRAM). However, such systems require refresh. In this paper, we examine the opportunity of minimizing on-chip memory power by intelligently refreshing a full-eDRAM cache hierarchy. We present Refrint, a simple approach to perform fine-grained, …
-
Hybrid caches: design and data management
… consumption. Among them, embedded Dynamic RAM (eDRAM) technology provides minimal area and leakage by design but reads are destructive and it is not as fast as SRAM. In this thesis, both SRAM and eDRAM technologies are mingled to take the advantatges that each of them o¿ers. First, they are …
-
Cache architectures based on heterogeneous technologies to deal with manufacturing errors
… brings alternative technologies such as eDRAM, which provides minimum area and power consumption.The main drawback of this memory technology is that reads are destructive and eDRAM cells work slower than SRAM ones. This thesis presents three main contributions regarding low-power caches …
-
Compute-in-memory designs for deep neural network and combinatorial optimization problems accelerators
… Among the prevalent semiconductor memories, eDRAM (embedded-DRAM) which integrates the DRAM bitcell monolithically along with high-performance logic transistors and interconnects can enable custom CIM designs by offering the densest embedded bitcell, low pJ/bit access energy, high-endurance, …
-
Statistical analysis techniques for logic and memory circuits.
… the yield. For memory circuits, embedded DRAM (eDRAM) is beginning to replace SRAM as the on-die cache choice in order to keep the scaling trend. Although techniques exist for statistical analysis of SRAM, detailed analysis of eDRAM has not been developed prior to this thesis. In this thesis, we …
-
Low-Power and Robust Level-Shifter with Contention Mitigation for Memory and Standalone Applications
… in today's technology. Memories such as eDRAM, are experiencing poor retention time because of challenges such as reference voltage variation, high transistor leakage, and low cell capacitance. It can be seen that we must consider not only the first order effects, but also the second …
-
Energy-Aware Data Movement In Non-Volatile Memory Hierarchies
… use of emerging technology alternatives such as eDRAM and Non-Volatile Memory (NVM) including STT-MRAM, PCM, and RRAM. The utilization of emerging technology in Last Level Cache (LLC) designs which occupies a signifcant fraction of total die area in Chip Multi Processors (CMPs) introduces new …
-
Architecture circuit co-design for machine learning and combinatorial optimization problems
… this dissertation presents a 3D embedded DRAM (eDRAM) TensorCore architecture to address the high computational density and memory bandwidth demands of modern machine learning workloads. Models like large language models and Neural Radiance Fields (NeRFs) process continuous, high-dimensional …
-
CMOS VLSI correlator design for radio-astronomical signal processing : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Engineering at Massey University, Auckland, New Zealand
… and prototyped in 130 nm that improve upon eDRAM by taking advantage of sequential access patterns. For the input-buffered architecture, a new scheme is proposed that decreases the usage of the input-buffer memory by a third by making use of multiple accumulators in every CMAC. Because …
-
In-Memory Computing Architecture for Deep Learning Acceleration
… field-effect transistors. The area-consuming eDRAM buffers are replaced by dense cross-point Spin Transfer Torque Magnetic RAM. I explored the design space and demonstrated that MARVEL can provide further improved power efficiency with increased number of integration layers. In the last piece …
-
Energy-efficient On-chip Deep Neural Network (DNN) Inference and Training with Emerging Non-volatile Memory Technologies
… and embedded dynamic random access memory (eDRAM), are not satisfactory in buffering a large amount of training data. Besides, the performance of eNVM-based in-memory training is limited by the high write energy and low write endurance of eNVM devices. In this thesis, I explored and …
-
Circuit architecture co-design for machine learning and 3D deep learning
… based on a monolithic 3D IGZO gain-cell eDRAM cube, which repurposes high-density memory structures as compute units. By exploiting 3D integration and an optimized inner/outer-product dataflow, this architecture improves compute density and energy efficiency for large-scale matrix …
-
Modeling, characterization, and control of the electrical-thermal interactions in advanced packages
The multi-chip integration in an advanced packaging has made the multi-physics interactions increasingly important. The objective of this research is to address the thermal coupling and the power density limitation of in-package systems through modeling, circuit emulation, and reduce thermal and …
-
Reducing cache misses due to frequent context switching using a cache context store
Computer system performance has been pushed further and further for decades, and hence the complexity of the designs has been increasing as well. This is true for both hardware and software. Problems at the interface of hardware and software are particularly interesting, as are solutions which …
-
Emerging semiconductor devices : design, modeling, and applications
… of IGZO-based two-transistor gain-cell eDRAM, accurately capturing geometry-dependent retention characteristics and highlighting the framework’s relevance for design–technology co-optimization. Collectively, the outcomes of this dissertation illustrate a coherent progression—from …