03 September 2008
Memory unit has become a major core component in most SoC designs, and thus reusable memory IP is crucial in speeding up the design process. In this thesis, we develop a low-power SRAM generator to reduce the design efforts by producing all the files required in traditional cell-based design flow. Several methods are used to reduce power consumption in the memory circuits, including hierarchical word-line architecture and block amplifiers. The SRAM generator can be extended to generate cache memory with mixed hard IP and soft IP where cache memory cells are hard IP while the cache controller is soft IP. Based on the SRAM generator, we can also generate some popular memory units such as register files, FIFO, LIFO, and delay elements used in many applications.
Hiles, David Roger.
No description available.
24 July 2013
In this thesis, I investigated the role of the hippocampus and other brain regions in long-term spatial memory. I used neuroimaging and behavioural techniques to compare spatial representations that are dependent on the hippocampus and those that rely on extra-hippocampal structures. In Experiment 1, I demonstrated that at least some spatial memories that are initially dependent on the hippocampus can become independent of it with time. This was done by looking at changes in brain activation as participants became familiar with a real-world environment over a year. Hippocampal activation was found during mental navigation tasks when participants were new to Toronto, but not after they had lived there for a year. This change was accompanied by an increase in activation in the posterior parahippocampal gyrus, lingual gyrus, cuneus, and superior temporal gyrus. In Experiment 2, I used neuroimaging to compare hippocampal involvement during the retrieval of coarse- and fine-grained spatial details, and episodic details associated with a familiar environment. I showed that hippocampally-mediated representations contain more fine-grained spatial details than extra-hippocampal spatial representations. Further, I demonstrated that fine-grained details become less dependent on the hippocampus with experience, but episodic details require the hippocampus throughout the lifetime of the memory. In Experiment 3, I provided behavioural evidence that the role of hippocampus in episodic memory associated with a familiar environment is crucially related to the degree of detail retrieved. Older adults were asked to recall walking routes from their daily lives and the number of details retrieved was correlated with tests sensitive to hippocampal function. Finally, I showed that the extra-hippocampal regions implicated in spatial memory depend on task demands. Basic navigation can be supported by the posterior parahippocampal gyrus, lingual gyrus, superior temporal gyrus, caudate and inferior frontal gyrus, independently of the hippocampus. When navigation requires fine-grained spatial details, additional regions including the precuneus and supramarginal gyrus will be recruited. When the task requires the retrieval of episodic details, the posterior cingulate, angular gyrus, and medial frontal lobes will be required, along with the hippocampus.
The assessment of very long-term memory using the questionnaire technique : a replication and extensionCzerwinski, Mary P. January 1983 (has links)
A replication and extension of a questionnaire technique for assessing very long-term memory is described. A 90-item questionnaire containing items from five different categories was administered to an Introductory Psychology class consisting of 90 females and 57 males. Results from each of the five categories revealed deviations from a linear trend in the recognition scores of items for a given year. Also, the results indicated that large proportions of the variance were due to the stimulus items used for that particular category. A post hoc test was included for two of the categories which tested the relationship between the number of references an item received in the New York Times Index with its recognition score on the questionnaire. A significant relationship was found and this evidence was used to support the hypothesis that the more available an item is to the general public, the more recognizable or distinctive it becomes. This distinctiveness causes the item to be resistant to memory decay and more recognizable, regardless of age.
Expectations modulate early but not late event-related potential correlates of successful recognition memoryKnight, Emily E. 07 November 2014 (has links)
Recognizing that an item has been previously encountered may not only depend on the strength of memory for that item, but also the expectation that the item will be remembered. Recent studies by O’Connor, et al (2010) and Jaeger et al (2013) revealed that a significant portion of the functional magnetic resonance imaging (fMRI) “retrieval success” effect (BOLD signal for correctly identified old items > new items) depends upon whether participants expect novel or familiar stimuli. The current study examined how expectancy modulates the event-related potential (ERP) retrieval success effect. We employed a typical recognition memory task with the addition of explicit cues indicating if upcoming memory probes were “likely old”, “likely new” (with 75% validity), or “unknown”. An electrophysiological response to the cue, primarily across frontal electrodes from 700-850ms after cue onset, predicted individual differences in cue- induced bias in memory judgments. Responses to memory probes were examined 300- 400ms and 500-700ms after probe onset, corresponding to time windows previously associated with ERP correlates of memory processing. Differences between old and new items were greatest from 300-400ms when preceded by “likely old” cues, overlapping with a component previously identified as tracking familiarity-based processing. In contrast, the 500-700ms time window, previously associated with recollection, revealed significant differences between old and new items that were not modulated by cue type. Overall this pattern of results shows that cue induced biases influence earlier (300-400ms after onset of a memory probe) but not later retrieval processing (500-700ms). / text
Leung, Wing-in, 梁穎妍
The present study aims at examining the relationship between specific memory recall and future event processing. Previous research suggested that the ability in generating specific future events was closely linked with recalling specific past memories. A total of 73 participants (40 in depressed group and 33 healthy control group) completed a specific memory recall and a future event generation task with emotional cues (positive and negative) and a neutral cue. The number of specific responses generated and the mean specific memory latency were recorded for analysis. Result not only replicated previous findings that the depressed group demonstrated more difficulties in generating specific responses, it also provided new insight in exploring the relationship between specific memory recall and future events processing. Firstly, the result indicated that the number of specific responses was fewer and the mean latency of specific responses was longer in generating future events than recalling specific memories. Secondly, the number and the mean latency of specific responses depended on different emotional loaded cues. Further discussion was focused on the valence effect on specific responses. Lastly, it was revealed that depressed patients had a particular difficulty in generating positive future events. Discussion was emphasized on the emotional loaded effect and hopelessness. Both theoretical and clinical implicated were suggested based on the new findings. / published_or_final_version / Clinical Psychology / Master / Master of Social Sciences
Pepper, Ross Laurence
Typescript. / Bibliography: leaves -151. / x, 151 l illus
Experimental studies in spelling to determine whether continuous repetition or repetition after an interval is more effective in memorizing,Eisenberg, John Linwood, January 1913 (has links)
Thesis (Ph. D.)--University of Pennsylvania, 1913.
Humpstone, Henry Judson,
Thesis (Ph. D.)--University of Pennsylvania, 1917. / Includes bibliographical references (p. 20).
Finkenbinder, Erwin Oliver,
Thesis (Ph. D.)--Clark University, 1913. / "Reprinted from the American journal of psychology XXV, 1914." "From the psychological laboratory of Clark university, pp. 32-81."
Page generated in 0.1005 seconds