Search

Zusammenfassung Von 4430 im Zeitraum von 2 Jahren untersuchten Huhnern wurden 276mal (= 6,2%) Vibrionen nachgewiesen. Damit sind die Vibrionenfunde in unserem Einzugsgebiet von 12,5% auf 6,2% zuruckgegangen. Die Vermehrung der Vibrionen war am erfolgreichsten in einer Stickstoffatmosphare bei einem CO2-Anteil von 20%. 14 von 20 Stammen waren in der Lage H2S zu bilden. Das serologische Verhalten von 54 mittels der KBR getesteten Stamme war sehr uneinheitlich. Obwohl 22 Stamme sich 2 serologischen Gruppen zuordnen liesen, erscheint das Verfahren zur Differenzierung von Huhnervibrionen nicht geeignet. Die karbolloslichen Antigene reagierten spezifischer als durch Hitze denaturierte Vibrionenantigene. Eintagskuken konnten nicht erfolgreich infiziert werden. Das steht in Ubereinstimmung mit der im Schrifttum vertretenen Ansicht, das fur die Vibrionenhepatitis des Huhnes mehrere Faktoren verantwortlich zu machen sind. Summary Antigenic properties of strains of vibrio isolated from chickens Of 4,430 chickens investigated over a period of two years, 276 (6.2%) were positive for vibrios. This means a fall in the incidence from 12.5% to 6.2%. Vibrios grew best in an atmosphere of 20% CO2 and 80% nitrogen. Of 20 strains, 14 were able to produce H2S. 54 strains were tested by the complement fixation test. The serological behaviour of these strains was very varied and although 22 strains fell into two serological groups, the method appeared unsuitable for differentiating chicken vibrios. Antigens soluble in phenol were more specific than antigens denatured by heat. Day-old chicks could not be successfully infected. This agrees with the view expressed in the literature that several factors are responsible for avian vibrionic hepatitis. Resume Proprietes antigeniques des souches de vibrions isolees chez les poules Sur 4430 poules examinees au cours d'une periode de deux ans, on met en evidence 276 fois des vibrions (6,2%). Les vibrions se trouvent donc en regression dans notre materiel d'analyse de 12,5% a 6,2%. La multiplication des vibrions reussit le mieux en atmosphere azotee avec 20% de CO2. Sur 20 souches, 14 sont capables de former du H2S. Le comportement serologique de 54 souches testees a la RFC manque completement d'unite. Bien que 22 souches puissent etre attribuees a deux groupes serologiques, ce procede ne parait pas indique pour la differenciation des vibrions de la poule. Les antigenes solubles dans le phenol reagissent de facon plus specifique que les antigenes de vibrions denatures a la chaleur. On ne parvient pas a infecter avec succes des poussins d'un jour. Ces observations concordent avec le point de vue soutenu dans la litterature, que plusieurs facteurs doivent etre consideres comme responsables de l'hepatite a vibrions chez la poule. Resumen Propiedades antigenicas de las estirpes aisladas de vibriones aviares De 4430 gallinas examinadas a lo largo de 2 anos, se apreciaron 276 veces (= 6,2%) vibriones. Con ello, los hallazgos de vibriones disminuyeron en la zona desde donde nos envian animales para proceder a su necropsia de 12,5% a 6,2%. La multiplicacion de los vibriones tuvo mayor exito en una atmosfera de nitrogeno con una proporcion de CO2 del 20%. 14 de 20 estirpes eran capaces de formar H2S. El comportamiento serologico de 54 cepas analizadas con ayuda de la prueba de fijacion del complemento era harto desigual. Aunque 22 estirpes se podian subordinar a 2 grupos serologicos, no resulta apropiada la tecnica para diferenciar los vibriones aviares. Los antigenos solubles en fenol reaccionaban con especificidad mayor que los antigenos vibrionicos desnaturalizados por el calor. Los pollitos de un dia de edad no se pudieron infectar con exito. Este hecho se halla en concordancia con el punto de vista propugnado en la bibliografia, de que para la hepatitis vibrionica de las gallinas se deben responsabilizar varios factores.

We present a new, simple algorithmic idea for the collective communication operations broadcast, reduction, and scan (prefix sums). The algorithms concurrently communicate over two binary trees which both span the entire network. By careful layout and communication scheduling, each tree communicates as efficiently as a single tree with exclusive use of the network. Our algorithms thus achieve up to twice the bandwidth of most previous algorithms. In particular, our approach beats all previous algorithms for reduction and scan. Experiments on clusters with Myrinet and InfiniBand interconnect show significant reductions in running time for all three operations sometimes even close to the best possible factor of two.

Modern computers have deepening memory hierarchieswith multiple levels of (partially shared) cachesand non-uniform memory access (NUMA). This makes itincreasingly difficult and important to schedule computationsin such a way that expensive memory accesses are avoided.In this paper we are choosing LU-decomposition for a case study since its use in the famous LINPACK benchmark means that highly tuned codes are already available. Our approach is to perform the very same computations as a leading implementation (PLASMA) but to schedule them in a more locality aware way. In particular, we better take into account when independent subtasks share the same input data and we explicitly address NUMA-effects coordinating memory layout and task scheduling. These measures lead to up to 36 % performance improvement compared to PLASMA.

We present an algorithm for matrix inversion that combines the practical requirement of an optimal number of arithmetic operations and the theoretical goal of a polylogarithmic critical path length. The algorithm reduces inversion to matrix multiplication. It uses Strassen's recursion scheme but on the critical path, it breaks the recursion early switching to an asymptotically inefficient yet fast use of Newton's method. We also show that the algorithm is numerically stable. Overall, we get a candidate for a massively parallel algorithm that scales to exascale systems even on relatively small inputs. Preliminary experiments on multicore machines give the surprising result that even on such moderately parallel machines the algorithm outperforms Intel's Math Kernel Library and that Strassen's algorithm seems to be numerically more stable than one might expect.

We give an efficient algorithm for solving the following scheduling problem to optimality: Assign n jobs to m processors such that they all meet a common deadline T and energy consumption is minimized by appropriately controlling the clock frequencies of the processors. Jobs are malleable, i.e., their amount of parallelism can be flexibly adapted. In contrast to previous work on energy efficient scheduling we allow more realistic energy consumption functions including a minimum and maximum clock frequency and a linear term in energy consumption. We need certain assumptions on the speedup function of the jobs that we show to apply for a large class of practically occurring functions.

We give an $O(n + \min\{n,m\} \log{m})$ work algorithm for scheduling $n$ tasks with flexible amount of parallelism on $m$ processors, provided the speedup functions of the tasks are concave. We give efficient parallelizations of the algorithm that run in polylogarithmic time. Previous algorithms were sequential and required quadratic work. This is in some sense a best-possible result since the problem is NP-hard for more general speedup functions.

We present a new, simple algorithmic idea for exploiting the capability for bidirectional communication present in many modern interconnects for the collective MPI operations broadcast, reduction and scan. Our algorithms achieve up to twice the bandwidth of most previous and commonly used algorithms. In particular, our algorithms for reduction and scan are the currently best known. Experiments on clusters with Myrinet and InfiniBand interconnects show significant reductions in running time for broadcast and reduction, for reduction even close to the best possible factor of two.

1. The two single slit sensilla on the pretarsus of the spider leg (Cupiennius salei Keys.) are vibration sensitive. Due to their arrangement they are also stimulated effectively by active downward movement of the pretarsus which may result from muscular activity directly or from an increase of the hemolymph pressure (Figs. 1, 2, 3). 2. The slits are not tuned to a restricted frequency range. Their threshold displacement was determined electrophysiologically from 0.01 Hz to 1 kHz and found to be higher than in the slits of the metatarsal organ by roughly two orders of magnitude. The threshold decreases only slightly up to about 40 Hz but much more steeply at higher frequencies. The minimal threshold value for displacement within the tested frequency range was found at 1 kHz and measured 2 × 10−5 cm. The minimal threshold value for acceleration found at 0.01 Hz was 0.3 × 10−9 cm/s2 (Fig. 4a). Thresholds vary with different preset tensions in the cuticle surrounding the slits (Fig. 5). 3. The shapes of the threshold curves are similar for dorsoventral and lateral displacement of the pretarsus (Fig. 4b). Values for lateral displacement are at most 3 dB higher than those for dorsoventral displacement. 4. As in other slit sensilla the frequency response of the pretarsal slits follows a power functiony(f) =fk (Fig. 7). The constantk varies between 0.39 and 0.44, reflecting receptor properties intermediate between those of a frequency-independent amplitude receptor (k = 0) and those of a velocity receptor (k = 1). Beyond 10 Hz the response of the receptor saturates. Phase locking between stimulus and response is observed up to about 100 Hz (stimuli 10 dB above threshold).

1. The first steps of the central nervous integration of sensory input from vibration receptors in the spiderCupiennius salei Keys, take place in the subesophageal ganglionic mass. We describe ventro-medially located neurons which respond to vibrations reaching the spider through the substrate (SVSN, SubstrateVibration Sensitive Neurons). 2. These neurons differ with respect to the convergence of their sensory input. Some neurons (SVSN type 1) only respond to the vibration of the leg corresponding to the ganglion recorded from. Other neurons can be effectively stimulated by vibrating any of the four ipsilateral legs (SVSN type 2) or either one of the two legs of the same segment (SVSN type 4). Still another type receives input from all eight legs. 3. According to their tuning curves (10 Hz–1 kHz) all vibration sensitive neurons in the subesophageal ganglionic mass so far studied behave like bandpass filters. Their best frequencies are in the frequency range of courtship and prey signals encountered by the spider in its natural habitat. Whereas the best frequencies of SVSN type 1 are found at low frequencies (ca. 80 Hz–100 Hz), medium (ca. 400 Hz), and high frequencies (ca. 800 Hz), respectively (Fig. 1), SVSN type 2, 3, and 4 are more broadly tuned to 80 Hz–200 Hz (Figs. 2, 3). 4. SVSN type 2 are more sensitive by about 12 dB to front leg stimulation than to hind leg stimulation (Fig. 2), and SVSN type 4 are more sensitive (by ca. 7 dB) to simultaneous vibration of both segmental legs than to the stimulation of either one of them alone (Fig. 3). 5. Threshold values (displacement of tarsus) at the best frequencies are in the range of 0.1 to 1 μm. They decrease by up to ca. 20 dB if band limited ‘noisy’ stimuli (bandwidth ca. 1/3 octave) are applied instead of sinusoidal vibration (Fig. 4). 6. The relation between stimulus amplitude and spike response is sigmoid. The dynamic ranges of the responses differ in the various neuron types. The low and high frequency SVSN type 1 have a dynamic range of more than 46 dB; in SVSN type 4 it increases if the two segmental legs are stimulated simultaneously instead of alone (Fig. 5).