Inheritance of synchronization and recovery properties inAvalon/C++

The authors describe their experience adapting inheritance mechanisms to a new application domain, reliable distributed systems. They give an overview of Avalon/C++, a programming language under development that allows programmers to `customize' the synchronization and fault-tolerance properties of data types by letting them inherit properties such as serializability and crash recovery from a library of basic types. The authors first describe the transaction model used to organize distributed computations and some relevant features of C++, and give an overview of the Avalon/C++ base hierarchy. They then describe in more detail each of the hierarchy's classes and some restrictions on their use that must be obeyed to preserve their semantic intent. An extended example illustrates a directory-type implementation that uses all three of the base classes. Related work is discussed     

Anti-Müllerian hormone in grazing dairy cows: Identification of factors affecting plasma concentration,relationship with phenotypic fertility,and genome-wide associations

The objectives of this study were to (1) characterize the distribution and variability of plasma anti-Müllerian hormone (AMH) concentration; (2) evaluate factors associated with phenotypic variation in plasma AMH; (3) examine the associations between categories of plasma AMH and reproductive outcomes [pregnancy to first artificial insemination (P/AI), and pregnancy rates within 21, 42, and 84 d after the mating start date (MSD)]; (4) estimate pedigree and genomic heritability for plasma AMH; and (5) identify and validate SNP associated with phenotypic variation in plasma AMH. Plasma AMH concentration (pg/mL) was determined from a blood sample collected (mean ± standard deviation) 10 ± 2 d after first insemination at detected estrus (IDE) in 2,628 first- and second-parity Irish dairy cows. Overall, plasma AMH had a positively skewed distribution with mean (± standard deviation), median, minimum, and maximum concentrations of 326 ± 231, 268, 15, and 2,863 pg/mL, respectively. Plasma AMH was greatest for Jersey, followed by Holstein × Jersey, Holstein × Norwegian Red, and Holstein cows (410, 332, 284, and 257 pg/mL, respectively). Second-parity cows had greater plasma AMH than first-parity cows (333 vs. 301 pg/mL, respectively). Samples collected at 7 and 8 d after first IDE had lesser plasma AMH than those collected on d 9, 10, 11, 12, and 13 after first IDE (291 and 297 vs. 317, 319, 331, 337, and 320 pg/mL). Plasma AMH was not associated with either body condition score at first IDE or the interval from calving to MSD. Cows were categorized into low (≤150 pg/mL; n = 526; lowest 20%), intermediate (>150 to ≤461 pg/mL; n = 1,576; intermediate 60%), and high AMH (>461 pg/mL; n = 526; highest 20%) groups based on plasma AMH, and associations with reproductive outcomes were tested. Cows with high and intermediate plasma AMH had 1.42- and 1.51-times-greater odds of becoming pregnant within 84 d after the MSD than those with low plasma AMH (90.3 and 90.8 vs. 86.8%, respectively); however, P/AI and pregnancy rate within 21 and 42 d after the MSD did not differ among AMH categories. Plasma AMH was moderately heritable (pedigree heritability of 0.40 ± 0.06 and genomic heritability of 0.45 ± 0.05), and 68 SNP across Bos taurus autosomes 7 and 11 were associated with phenotypic variation in plasma AMH. Out of 68 SNP, 42 were located in a single quantitative trait locus on Bos taurus autosome 11 that harbored 6 previously identified candidate genes (NR5A1, HSPA5, CRB2, DENND1A, NDUFA8, and PTGS) linked to fertility-related phenotypes in dairy cows.     

Effect of equine chorionic gonadotropin treatment during a progesterone-based timed artificial insemination program on reproductive performance in seasonal-calving lactating dairy cows

The aim of this study was to investigate the effect of progesterone (P4)-based timed artificial insemination (TAI) programs on fertility in seasonal-calving, pasture-based dairy herds. A total of 1,421 lactating dairy cows on 4 spring-calving farms were stratified based on days in milk (DIM) and parity and randomly allocated to 1 of 3 treatments: (1) control: no hormonal treatment; cows inseminated at detected estrus; (2) P4-Ovsynch: cows received a 7-d P4-releasing intravaginal device (PRID Delta; CEVA Santé Animale, Libourne, France) with 100 μg of a gonadotropin-releasing hormone (GnRH) analog (Ovarelin; CEVA Santé Animale) at PRID insertion, a 25-mg injection of PGF_2α (Enzaprost; CEVA Santé Animale) at PRID removal, GnRH at 56 h after device removal and TAI 16 h later; (3) P4-Ovsynch+eCG: the same as P4-Ovsynch, but cows received 500 IU of equine chorionic gonadotropin (eCG; Syncrostim; CEVA Santé Animale) at PRID removal. At 10 d before mating start date (MSD), all cows that were ≥35 DIM were examined by transrectal ultrasound to assess presence or absence of a corpus luteum; body condition score (BCS) was also recorded. Pregnancy diagnosis was performed by transrectal ultrasonography 30 to 35 d after insemination. Overall pregnancy/AI (P/AI) was not different between groups (50.9, 49.8, and 46.3% for control, P4-Ovsynch, and P4-Ovsynch+eCG, respectively) but the 21-d pregnancy rate was increased by the use of synchronization (35.0, 51.7, and 47.2%, respectively). Compared with the control group, synchronization significantly reduced the interval from MSD to conception (34.6, 23.0, and 26.5 d, respectively) and consequently reduced the average days open (98.0, 86.0, and 89.0 d). Across all treatment groups, DIM at the start of synchronization affected P/AI (42.3, 49.5, and 53.9% for <60, 60–80, and >80 DIM, respectively), but neither parity (46.5, 50.4, and 48.4% for parity 1, 2, and ≥3, respectively) nor BCS (44.0, 49.4, and 58.6% for ≤2.50, 2.75–3.25, and ≥3.50, respectively) affected the likelihood of P/AI. Two-way interactions between treatment and DIM, parity, or BCS were not detected. In conclusion, the use of TAI accelerated pregnancy establishment in cows in a pasture-based system by reducing days open, but eCG administration at PRID removal did not affect P/AI.     

Hardware Transactional Memory Exploration in Coherence-Free Many-Core Architectures

High-end embedded systems, like their general-purpose counterparts, are turning to many-core cluster-based shared-memory architectures that provide a shared memory abstraction subject to non-uniform memory access costs. In order to keep the cores and memory hierarchy simple, many-core embedded systems tend to employ simple, scratchpad-like memories, rather than hardware managed caches that require some form of cache coherence management. These “coherence-free” systems still require some means to synchronize memory accesses and guarantee memory consistency. Conventional lock-based approaches may be employed to accomplish the synchronization, but may lead to both usability and performance issues. Instead, speculative synchronization, such as hardware transactional memory, may be a more attractive approach. However, hardware speculative techniques traditionally rely on the underlying cache-coherence protocol to synchronize memory accesses among the cores. The lack of a cache-coherence protocol adds new challenges in the design of hardware speculative support. In this article, we present a new scheme for hardware transactional memory (HTM) support within a cluster-based, many-core embedded system that lacks an underlying cache-coherence protocol. We propose two alternative data versioning implementations for the HTM support, Full-Mirroring and Distributed Logging and we conduct a performance comparison between them. To the best of our knowledge, these are the first designs for speculative synchronization for this type of architecture. Through a set of benchmark experiments using our simulation platform, we show that our designs can achieve significant performance improvements over traditional lock-based schemes.     

Sorting and Counting Networks of Arbitrary Width and Small Depth

We present the first construction for sorting and counting networks of arbitrary width that requires both small depth and small constant factors in the depth expression. Let w be the product w = p ₀ ⋅ p ₁ ⋅s p _n-1 , whose factors are not necessarily prime. We present a novel network construction of width w and depth O(n ² ) = O(log ² w) , using comparators (or balancers) of width less than or equal to max(p _i ) . This construction is practical in the sense that the asymptotic notation does not hide any large constants. An interesting aspect of this construction is that it establishes a family of sorting and counting networks of width w , one for each distinct factorization of w . A factorization in which max(p _i ) is large and n is small yields a network that trades small depth for large comparators (or balancers), and a factorization where max(p _i ) is small and n is large makes the opposite tradeoff. Received June 18, 2001. Online publication October 30, 2001.     

Effects of restrictions and prices of N,P and K on fertiliser inputs and yield deficits of sugar beet

Although excessive application of fertiliser has been commonplace, optimisation, as well as more limited use, is increasingly topical because of environmental and economic considerations that provide an impetus for either direct or price-induced restriction of fertiliser use. In this context, optimum N, P and K inputs, yields of sugar and yield deficits were calculated for sugar beet for current nutrient prices, multiples of 1.5, 2, 3 and 5 times current prices and for a range from 0% to 150% of the current optimum for each of four soil-fertility index categories. In terms of the optimum input, N was more impervious to price increase than P or K. For example, on low-fertility index 1 soils, reduction of inputs to circa 75% of current optimum required increases of 200% for N, 50% for P and 100% for K. Increases of 20–30%, representative of fertiliser taxes, had little impact. Variation in product price had a greater effect than nutrient price, as indicated by the exponential relationship between their proportional changes for similar effects on the price ratio, and the combined effect of small changes in each was appreciable.All nutrient-price increases induced yield deficits in the order P > K > N. The aggregate financial effect on yield deficit and net nutrient cost, i.e. on gross margin, was in the order N > K > P for very large price increases, with little difference between nutrients for increases of 100% or less. Mean effects of 100% price increases were yield deficits (%) of 0.3 to 0.4 for N, 0.7 to 2.3 for P and 0.4 to 1.3 for K, depending on index category, and reductions in gross margins (%) of 5.1 to 5.4 at index 1 and 0.7 to 1.6 at index 4 depending on the nutrient. Direct input restriction to 75% or 50% of optimum had negligible financial effects, with the exception of the 50% restriction for N. Voluntary restriction of fertiliser inputs for sugar beet, therefore, would be both environmentally and financially efficient compared with price-induced restriction at the level of a fertiliser tax.Since distribution of yield deficits was right-skewed, a two-parameter gamma distribution was used to determine the probability of exceeding deficits of 5% and 3%. Only extreme price increases or input restrictions, or excessive use in the case of N, resulted in deficit frequencies of significant proportions; comparatively, P and K were more sensitive to price increase and N more sensitive to direct restriction. Complete input restriction on high-fertility, index 4, soils breached the acceptable tolerance of deficits >5% for N and K but not for P. For the latter, the results implied a possible conflict between environmental good-practice and the achievement of nutrient reserves required for highest yield.     

The topology of distributed adversaries

Roughly speaking, a simplicial complex is shellable if it can be constructed by gluing a sequence of n-simplexes to one another along $(n-1)$ ( n ? 1 ) -faces only. Shellable complexes have been widely studied because they have nice combinatorial properties. It turns out that several standard models of concurrent computation can be constructed from shellable complexes. We consider adversarial schedulers in the synchronous, asynchronous, and semi-synchronous message-passing models, as well as asynchronous shared memory. We show how to exploit their common shellability structure to derive new and remarkably succinct tight (or nearly so) lower bounds on connectivity of protocol complexes and hence on solutions to the $k$ k -set agreement task in these models. Earlier versions of material in this article appeared in the 2010 ACM Symposium on Principles of Distributed Computing (Herlihy and Rajsbaum 2010), and the International Conference on Distributed Computing (Herlihy and Rajsbaum 2010, doi:10.1145/1835698.1835724).