Reducing main memory access latency through SDRAM address mapping techniques and access reordering mechanisms

As the performance gap between microprocessors and memory continues to increase, main memory accesses result in long latencies which become a factor limiting system performance. Previous studies show that main memory access streams contain significant localities and SDRAM devices provide parallelism through multiple banks and channels. These locality and parallelism have not been exploited thoroughly by conventional memory controllers. In this thesis, SDRAM address mapping techniques and memory access reordering mechanisms are studied and applied to memory controller design with the goal of reducing observed main memory access latency. The proposed bit-reversal address mapping attempts to distribute main memory accesses evenly in the SDRAM address space to enable bank parallelism. As memory accesses to unique banks are interleaved, the access latencies are partially hidden and therefore reduced. With the consideration of cache conflict misses, bit-reversal address mapping is able to direct potential row conflicts to different banks, further improving the performance. The proposed burst scheduling is a novel access reordering mechanism, which creates bursts by clustering accesses directed to the same rows of the same banks. Subjected to a threshold, reads are allowed to preempt writes and qualified writes are piggybacked at the end of the bursts. A sophisticated access scheduler selects accesses based on priorities and interleaves accesses to maximize the SDRAM data bus utilization. Consequentially burst scheduling reduces row conflict rate, increasing and exploiting the available row locality. Using a revised SimpleScalar and M5 simulator, both techniques are evaluated and compared with existing academic and industrial solutions. With SPEC CPU2000 benchmarks, bit-reversal reduces the execution time by 14% on average over traditional page interleaving address mapping. Burst scheduling also achieves a 15% reduction in execution time over conventional bank in order scheduling. Working constructively together, bit-reversal and burst scheduling successfully achieve a 19% speedup across simulated benchmarks.

Comparisons between OMI SO2 data and ground-based SO2 measurements at Turrialba volcano

Turrialba is one of the largest and most active stratovolcanoes in the Central Cordillera of Costa Rica and an excellent target for validation of satellite data using ground based measurements due to its high elevation, relative ease of access, and persistent elevated SO2 degassing. The Ozone Monitoring Instrument (OMI) aboard the Aura satellite makes daily global observations of atmospheric trace gases and it is used in this investigation to obtain volcanic SO2 retrievals in the Turrialba volcanic plume. We present and evaluate the relative accuracy of two OMI SO2 data analysis procedures, the automatic Band Residual Index (BRI) technique and the manual Normalized Cloud-mass (NCM) method. We find a linear correlation and good quantitative agreement between SO2 burdens derived from the BRI and NCM techniques, with an improved correlation when wet season data are excluded. We also present the first comparisons between volcanic SO2 emission rates obtained from ground-based mini-DOAS measurements at Turrialba and three new OMI SO2 data analysis techniques: the MODIS smoke estimation, OMI SO2 lifetime, and OMI SO2 transect techniques. A robust validation of OMI SO2 retrievals was made, with both qualitative and quantitative agreements under specific atmospheric conditions, proving the utility of satellite measurements for estimating accurate SO2 emission rates and monitoring passively degassing volcanoes.

WATER ACCESS AND MAINTENANCE IN KARONGA, MALAWI

This report is a case study of how Mwangalala community accesses water and how that access is maintained. Mwangalala community is located in the northern tip of Karonga district in Malawi, Africa. The case study evaluates how close the community is to meeting target 10 of the Millennium Development Goals, sustainable access to safe drinking water, and evaluates the current water system through Human Centered Design’s criteria of desirability, feasibility, and viability. It also makes recommendations to improve water security in Mwangalala community. Data was collected through two years of immersive observation, interviews with 30 families, and observing two wells on three separate occasions. The 30 interviews provided a sample size of over 10% of the community’s population. Participants were initially self-selected and then invited to participate in the research. I walked along community pathways and accepted invitations to join casual conversations in family compounds. After conversing I asked the family members if they would be willing to participate in my research by talking with me about water. Data collected from the interviews and the observations of two wells were compared and analyzed for common themes. Shallow wells or open wells represented the primary water source for 93% of interview participants. Boreholes were also present in the community, but produced unpalatable water due to high concentrations of dissolved iron and were not used as primary water sources. During observations 75% of community members who used the shallow well, primarily used for consumptive uses like cooking or dinking, were females. Boreholes were primarily used for non-consumptive uses such as watering crops or bathing and 77% of the users were male. Shallow wells could remain in disrepair for two months because the repairman was a volunteer, who was not compensated for the skilled labor required to repair the wells. Community members thought the maintenance fee went towards his salary, so did not compensate the repairman when he performed work. This miscommunication provided no incentive for the repairman to make well repairs a priority, and left community members frustrated with untimely repairs. Shallow wells with functional pumps failed to provide water when the water table levels drop during dry season, forcing community members to seek secondary or tertiary water sources. Open wells, converted from shallow wells after community members did not pay for repairs to the pump, represented 44% of the wells originally installed with Mark V hand pumps. These wells whose pumps were not repaired were located in fields and one beside a church. The functional wells were all located on school grounds or in family compounds, where responsibility for the well’s maintenance is clearly defined. Mwangalala community fails to meet Millennium Development goals because the wells used by the community do not provide sustainable access to safe drinking water. Open wells, used by half the participants in the study, lack a top covering to prevent contamination from debris and wildlife. Shallow well repair times are unsustainable, taking longer than two weeks to be repaired, primarily because the repair persons are expected to provide skilled labor to repair the wells without compensation. Improving water security for Mwangalala can be achieved by improving repair times on shallow wells and making water from boreholes palatable. There are no incentives for a volunteer repair person to fix wells in a timely manner. Repair times can be improved by reducing the number of wells a repair person is responsible for and compensating the person for the skilled labor provided. Water security would be further improved by removing iron particulates from borehole water, thus rendering it palatable. This is possible through point of use filtration utilizing ceramic candles; this would make pumped water available year-round.