Search for: All records

Description / Abstract: In order to effectively provide INaaS (Inference-as-a-Service) for AI applications in resource-limited cloud environments, two major challenges must be overcome: achieving low latency and providing multi-tenancy. This paper presents EIF (Efficient INaaS Framework), which uses a heterogeneous CPU-FPGA architecture to provide three methods to address these challenges (1) spatial multiplexing via software-hardware co-design virtualization techniques, (2) temporal multiplexing that exploits the sparsity of neural-net models, and (3) streaming-mode inference which overlaps data transfer and computation. The prototype EIF is implemented on an Intel PAC (shared-memory CPU-FPGA) platform. For evaluation, 12 types of DNN models were used as benchmarks, with different size and sparsity. Based on these experiments, we show that in EIF, the temporal multiplexing technique can improve the user density of an AI Accelerator Unit from 2$$\times$$ to 6$$\times$$, with marginal performance degradation. In the prototype system, the spatial multiplexing technique supports eight AI Accelerators Unit on one FPGA. By using a streaming mode based on a Mediated Pass-Through architecture, EIF can overcome the FPGA on-chip memory limitation to improve multi-tenancy and optimize the latency of INaaS. To further enhance INaaS, EIF utilizes the MapReduce function to provide a more flexible QoS. Together with the temporal/spatial multiplexing techniques, EIF can support 48 users simultaneously on a single FPGA board in our prototype system. In all tested benchmarks, cold-start latency accounts for only approximately 5\% of the total response time. 

Abstract Charge‐programmed 3D printing enables the fabrication of 3D electronics with lightweight and high precision via selective patterning of metals. This selective metal deposition is catalyzed by Pd nanoparticles that are specifically immobilized onto the charged surface and promises to fabricate a myriad of complex electronic devices with self‐sensing, actuation, and structural elements assembled in a designed 3D layout. However, the achievable property space and the material‐performance correlation of the charge‐programmed printing remain unexplored. Herein, a series of photo‐curable resins are designed for unveiling how the charge and crosslink densities synergistically impact the nanocatalyst‐guided selective deposition in catalytic efficiency and properties of the 3D printed charge‐programmed architectures, leading to high‐quality 3D patterning of solid and liquid metals. The findings offer a wide tunability of the structural properties of the printed electronics, ranging from stiff to extreme flexibility. Capitalizing on these results, the printing and successful application of an ultralight‐weight and deployable 3D multi‐layer antenna system operating at an ultrahigh‐frequency of 19 GHz are demonstrated. 

Fertilizer utilization is critical for global food security. This study examines the occurrence of trace elements (TEs) and Sr isotope (87Sr/86Sr) variations in phosphate rocks and mineral fertilizers from a sample collection representative of major phosphate producing countries. We show high concentrations of several TEs in phosphate rocks (n=76) and their selective enrichment in phosphate fertilizers (n=40) of specific origin. Consistent with the concentrations in parent phosphate rocks, phosphate fertilizers from the U.S. and Middle East have substantially higher concentrations of U, Cd, Cr, V, and Mo than in fertilizers from China and India. Yet, fertilizers from China and India generally have higher concentrations of As. The 87Sr/86Sr in phosphate fertilizers directly mimic the composition of their source phosphate rocks, with distinctive higher ratios in fertilizers from China and India (0.70955–0.71939) relative to phosphate fertilizers from U.S. and Middle East (0.70748–0.70888). Potash fertilizers have lower Sr and TEs and higher 87Sr/86Sr (0.72017–0.79016), causing higher 87Sr/86Sr in mixed NPK-fertilizers. Selective extraction (Mehlich III) of soils from an experimental agricultural site shows relative enrichment of potentially plant-available P, Sr, and TEs in topsoil, which is associated with Sr isotope variation towards the 87Sr/86Sr of the local utilized phosphate fertilizer. 

Phosphate fertilizers may contain elevated concentrations of toxic metals and metalloids and therefore, their excessive application can result in the accumulation of both phosphorus (P) and metal(loid)s in agricultural soils. This study aims to investigate the occurrence, distribution, and potential plant-availability of metal(loid)s originating from phosphate fertilizer in a long-term experimental field at the Tidewater Research Station in North Carolina, where topsoil (10-20 cm deep) and subsoil (up to 150 cm deep) samples were collected from five plots with consistent and individually different application rates of P-fertilizer since 1966. We conducted systematic analyses of P and metal(loid)s in bulk soils, in the plant available fraction, and in four sequentially extracted soil fractions (exchangeable, reducible, oxidizable, and residual). The results show that P content in topsoils were directly associated with the rate of P-fertilizer application (=1, p<0.05). Furthermore, P concentrations were highly correlated with concentrations of Cd, U, Cr, V, and As in the bulk topsoil (>0.58, p<0.05), as well as the potential plant-available fraction (>0.67, p<0.01), indicating the accumulation of the fertilizer-derived toxic metal(loid)s in the topsoil. Significant correlations (p<0.001) of metal(loid)s concentrations between the bulk soil and the potential plant-available fraction raises the possibility that P-fertilizer application could increase the accumulation of toxic metal(loid)s in plants, which could increase human exposure. Results from sequential leaching experiments revealed that large portions of the trace elements, in particular Cd, occur in the soluble (exchangeable and reducing) fractions of topsoil with higher P-fertilizer input, whereas the levels of redox-sensitive elements (As, V, U, Cr) were higher in the reducible and oxidizable fractions of the soils. Overall, the data presented in this study demonstrate the effect of long-term P-fertilizer application on the occurrence and accumulation of a wide range of toxic metal(loid)s in agricultural topsoil. 

Abstract Topologically ordered phases of matter elude Landau’s symmetry-breaking theory, featuring a variety of intriguing properties such as long-range entanglement and intrinsic robustness against local perturbations. Their extension to periodically driven systems gives rise to exotic new phenomena that are forbidden in thermal equilibrium. Here, we report the observation of signatures of such a phenomenon—a prethermal topologically ordered time crystal—with programmable superconducting qubits arranged on a square lattice. By periodically driving the superconducting qubits with a surface code Hamiltonian, we observe discrete time-translation symmetry breaking dynamics that is only manifested in the subharmonic temporal response of nonlocal logical operators. We further connect the observed dynamics to the underlying topological order by measuring a nonzero topological entanglement entropy and studying its subsequent dynamics. Our results demonstrate the potential to explore exotic topologically ordered nonequilibrium phases of matter with noisy intermediate-scale quantum processors. 

Phosphate rock bears both geologically and environmentally significant information. Rare earth elements and yttrium (i.e., REY) characteristics have been commonly used for reconstructing the redox conditions of depositional environments of and the effects of post-depositional diagenetic alteration on phosphate rock. In addition, phosphate rock is typically enriched in a range of trace elements such as uranium (U) and cadium (Cd) that can be dispersed as contaminants into the environment with phosphate mining and phosphate fertilizer application. Here we report the lead (Pb) isotope compositions combined with Pb and REY concentrations of both global sedimentary and igneous phosphate rocks, aiming to evalute the geological origin of phosphate rocks over time and the potential of using them for environmental tracing. Phopshate rocks samples analyzed in this study were sourced from major economic phosphate deposits in the world, including China, Southern Tethys (e.g., Morocco, Tunisia, Israel), the U.S., India, South Africa and Russia. Our results show a wide range of 208Pb/204Pb (35.70 to 60.58), 207Pb/204Pb (15.20 to 18.25), and 206Pb/204Pb (16.369 to 71.806) ratios in phosphate rocks, with sedimentary phosphate rocks being significantly more radiogenic than igneous rocks. The majority of the sedimentray phosphate rocks show a notable isotopic overprinting by non-radiogenic terrestrial Pb, except for those from Israel and Morocco that have the most radiogenic Pb isotope compositions. Correspondingly, phosphate rocks with more radiogenic Pb isotope ratios show relatively pristine seawater REY features, likely suggesting their preservation of the original oxic seawater conditions and/or minimal diagenetic alteration. In contrast, phosphate rocks with less radiogenic Pb isotope compositions show REY indications for more anoxic seawater redox conditions and/or greater diagenetic alteration. We further evaluate the potential utility of Pb isotopes for tracing the associated contamination with phosphate rock mining and fertilizer application in the environment. In most cases, the radiogenic Pb isotope composition of phosphate rocks and corresponding P-fertilizers is distinctive from both natural crustal Pb and major anthropogenic Pb sources (e.g., Pb ore deposits and pesticides), which provides a great advantage for applying Pb isotopes as environmental tracers for metal(loid) contamination from phosphate sources. The combination of Pb isotope ratios and REY proxies could further constrain the Pb source discrimination. Overall, this study provides new Pb isotopic and REY geochemical data on global phosphate rocks and fertilizers, which lays the groundwork for future regional and local studies on both their geological and environmental implications. 

Multiparameter Franck–Condon analyses of absorption spectra of Y6 in dilute solutions reveals that Y6 exhibits a high conformation uniformity and the smallest intra-molecular reorganization energy among the materials studied. 

Despite extensive research and technology to reduce the atmospheric emission of Pb from burning coal for power generation, minimal attention has been paid to Pb associated with coal ash disposal in the environment. This study investigates the isotopic signatures and output rates of Pb in fly ash disposal in China, India, and the United States. Pairwise comparison between feed coal and fly ash samples collected from coal-fired power plants from each country shows that the Pb isotope composition of fly ash largely resembles that of feed coal, and its isotopic distinction allows for tracing the release of Pb from coal fly ash into the environment. Between 2000 and 2020, approx. 236, 56, and 46 Gg Pb from fly ash have been disposed in China, India, and the U.S., respectively, posing a significant environmental burden. A Bayesian Pb isotope mixing model shows that during the past 40 to 70 years, coal fly ash has contributed significantly higher Pb (∼26%) than leaded gasoline (∼7%) to Pb accumulation in the sediments of five freshwater lakes in North Carolina, U.S.A. This implies that the release of disposed coal fly ash Pb at local and regional scales can outweigh that of other anthropogenic Pb sources.