Search

The dissertation investigates redundant communication between servers for large-scaleweb and cache requests and redundant data movement between accelerators for compute-intensiveapplications. Redundancy is an impending and critical issue for data centers designed forhardware accelerators and disaggregated resources. The dissertation makes the following threecontributions to address this. The first contribution of the dissertation is Daronpon. Daronpondynamically load-balances and reroutes large-scale requests of web and cache applications on amicrosecond timescale. Daronpon prevents these requests, stranded on busy servers with networkcongestion and long queuing delays, from being processed. Daronpon shows improvement invarious service time characterizations of different applications. The second contribution of thedissertation is Fianchetto. Fianchetto acts as a compute-enabled bypass for inter-acceleratorcommunication. Fianchetto accelerates the data restructuring needed between accelerators andsaves the data movement between accelerators and CPUs for compute-intensive applications.Fianchetto shows improvement in a series of benchmarks involving different application domains.The third contribution of the dissertation is Aurelia. Aurelia leverages the emerging interconnectof CXL to investigate the design of a scalable fabric for accelerators and fabric-attached memoryexpansion. Aurelia improves routing and transport based on the current specification of CXL andshows performance improvement on machine learning and key-value store applications

This study investigated the concentration of the pesticide residues found in Fragaria and Myrica rubra sold in the city of Hangzhou, People's Republic of China. From an analysis of 151 (77 Fragaria and 74 M. rubra) samples using gas chromatography-tandem mass spectrometry (GC-MS/MS) and liquid chromatography-tandem mass spectrometry (LC-MS/MS), 41 pesticide residues were found to be present. Of the 41 residues, 14 were found using GC-MS/MS and 27 were found using LC-MS/MS. Of the 151 samples, 10 (13.0%) of the 77 Fragaria samples and 5 (6.8%) of the 74 M. rubra samples were found to contain a specific pesticide residue, and only 4 Fragaria samples and 2 M. rubra samples were found to be without pesticide residue. In addition, 18 of the 41 pesticides were not detected in either Fragaria or M. rubra samples. Of the 41 residues, 10 were detected in Fragaria samples and 20 were detected in M. rubra samples. In Fragaria, procymidone was the most commonly detected residue, with a detection rate of 88.3%, followed by prochloraz, with a detection rate of 53.2%. In M. rubra, prochloraz was the most commonly detected residue, with a detection rate of 71.6%, followed by carbendazim, with a detection rate of 68.9%. The pesticide residues in some samples exceeded the maximum residue limit set in China. The limit of dimethomorph was exceeded in three of the Fragaria samples, and that of dichlorvos was exceeded in two of the M. rubra samples.

While (1) serverless computing is emerging as a popular form of cloud execution, datacenters are going through major changes: (2) storage dissaggregation in the system infrastructure level and (3) integration of domain-specific accelerators in the hardware level. Each of these three trends individually provide significant benefits; however, when combined the benefits diminish. Specifically, the paper makes the key observation that for serverless functions, the overhead of accessing dissaggregated persistent storage overshadows the gains from accelerators. Therefore, to benefit from all these trends in conjunction, we propose Domain-Specific Computational Storage for Serverless (DSCS-Serverless). This idea contributes a serverless model that leverages a programmable accelerator within computational storage to conjugate the benefits of acceleration and storage disaggregation simultaneously. Our results with eight applications shows that integrating a comparatively small accelerator within the storage (DSCS-Serverless) that fits within its power constrains (15 Watts), significantly outperforms a traditional disaggregated system that utilizes the NVIDIA RTX 2080 Ti GPU (250 Watts). Further, the work highlights that disaggregation, serverless model, and the limited power budget for computation in storage require a different design than the conventional practices of integrating microprocessors and FPGAs. This insight is in contrast with current practices of designing computational storage that are yet to address the challenges associated with the shifts in datacenters. In comparison with two such conventional designs that either use quad-core ARM A57 or a Xilinx FPGA, DSCS-Serverless provides 3.7x and 1.7x end-to-end application speedup, 4.3x and 1.9x energy reduction, and 3.2x and 2.3x higher cost efficiency, respectively.

AIM: To evaluate the correlation between Demodex infestation and keratitis, and to assess demodicosis using a simple approach. METHODS: A modified slit lamp illumination (at 40× magnification) was used to observe Demodex tails in 40 patients with refractory keratitis and 80 healthy controls. Bacterial smear and culture of the conjunctival sac and corneal lesion were performed to identify the pathogen. Tea tree oil ointment (TTOO) was added as a Demodex killing agent for lid scrubs to the treatment when Demodex infestation was confirmed. RESULTS: Demodex tails were found in all patients compared to 42/80 of the controls (P