Uniform Latency Research Articles

Spawning performance of Clarias gariepinus (Burchell, 1822) induced with ethanol preserved and fresh catfish pituitary extract.

This study was conducted to determine the efficacy of preserved and fresh pituitary extract of Clarias gariepinus (Burchell, 1822) to induced spawning in the same species. Growth performance of fry was also monitored for 15 days to determine the possible effect of natural hormone treatment on the fry. Pituitary glands (PG) were obtained from 12 male broodstocks, six extracted PG were preserved in 96% ethanol 24 h before injection, while the other were extracted and used in their fresh state. The PG solutions (1 g ml-1 of saline water) were injected at 1 ml kg-1 of the female broodstock. Ovaprim® (a synthetic hormone) was used as the control and administered at a rate of 0.5 ml kg -1 of fish. After a uniform latency period of 9 h 30 min, fish were striped, fertilized, incubated and the performance evaluated. Results obtained revealed better hatching performance using Ovaprim® synthetic hormone (64.52%), however, preserved PG gave better hatchability (59.74%) than fresh PG extract (51.39%). After 15 days of feeding ad libitum with shell-free Artemia cysts, the growth of fry spawn with preserved PG was comparable with that of the control, while least performance was observed using the fresh PG. It was concluded that preservation of PG in ethanol 24 h before injection had a positive effect on breeding performance and could be exploited in the commercial production of C. gariepinus fingerlings.

Circuit design of Clos-based on-chip interconnection networks

Single-hop non-blocking networks have the advantage of providing uniform latency and throughput, which is important for cache-coherent network-on-chip systems. This paper focuses on high performance circuit designs of multi-stage non-blocking networks as alternatives to crossbars. Existing work shows that Benes networks have much lower transistor count and smaller circuit area but longer delay than crossbars. To reduce the timing delay, we propose to design the Clos network built with larger size switches. Using less than half number of stages than the Benes network, the Clos network with 4×4 switches can significantly reduce the timing delay. The circuit designs of both Benes and Clos networks in different sizes are conducted considering two types of implementation of the configurable switch: with N-type metal-oxide-semiconductor logic (NMOS) transistors only and full transmission gates (TGs). The layout and simulation results under 45nm technology show that the TG-based implementation demonstrates much better signal integrity than its counterpart. Clos networks achieve average 60% lower timing delay than Benes networks with even smaller area and power consumption.

Design and Evaluation of a Flexible-Bandwidth OFDM-Based Intra-Data Center Interconnect

Data center networks are facing growing challenges to deliver higher bandwidth efficiency, lower latency, better flexibility, and lower cost. Various optical interconnect schemes have been proposed to take advantage of the high bandwidth capacity and low power consumption offered by optical switching. However, these schemes cannot offer flexible bandwidth sharing due to the large granularity in optical circuit switching, and they require costly optical components. In this paper, we introduce a novel data center network architecture based on cyclic arrayed waveguide grating device and multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing technology with parallel signal detection (PSD). This architecture offers flexible bandwidth resource sharing at fine granularity. Other features include high-speed switching, low and uniform latency, and the ability to change the data rates dynamically. By eliminating costly optical components and keeping the core optical router passive and static, the power consumption, hardware cost, and operation cost are reduced. The fine granularity bandwidth sharing and MIMO switching through PSD are verified experimentally. We also propose and evaluate efficient subcarrier allocation schemes to achieve high bandwidth utilization. Finally, we present the implementation of an efficient scheduler for the bandwidth allocation of the proposed scheme.

High-Bandwidth Network Memory System Through Virtual Pipelines

As network bandwidth increases, designing an effective memory system for network processors becomes a significant challenge. The size of the routing tables, the complexity of the packet classification rules, and the amount of packet buffering required all continue to grow at a staggering rate. Simply relying on large, fast SRAMs alone is not likely to be scalable or cost-effective. Instead, trends point to the use of low-cost commodity DRAM devices as a means to deliver the worst-case memory performance that network data-plane algorithms demand. While DRAMs can deliver a great deal of throughput, the problem is that memory banking significantly complicates the worst-case analysis, and specialized algorithms are needed to ensure that specific types of access patterns are conflict-free. We introduce virtually pipelined memory, an architectural technique that efficiently supports high bandwidth, uniform latency memory accesses, and high-confidence throughput even under adversarial conditions. Virtual pipelining provides a simple-to-analyze programming model of a deep pipeline (deterministic latencies) with a completely different physical implementation (a memory system with banks and probabilistic mapping). This allows designers to effectively decouple the analysis of their algorithms and data structures from the analysis of the memory buses and banks. Unlike specialized hardware customized for a specific data-plane algorithm, our system makes no assumption about the memory access patterns. We present a mathematical argument for our system's ability to provably provide bandwidth with high confidence and demonstrate its functionality and area overhead through a synthesizable design. We further show that, even though our scheme is general purpose to support new applications such as packet reassembly, it outperforms the state-of-the-art in specialized packet buffering architectures.

Eye movements during list execution reveal no planning in monkeys (Macaca fascicularis).

When trained on a serial-order task to respond to 5 simultaneously displayed stimuli, monkeys (Macaca fascicularis) show an elevated latency to respond to the first item (A) or the first 2 items (A and B) followed by short and uniform latencies to respond to the remaining items. This pattern has been interpreted as evidence of planning, with the long initial latency reflecting the animal planning their responses, and the subsequent short and uniform latencies reflecting the execution of the planned responses. An analysis of eye movements, however, revealed no evidence of planning. The elevated latency to item A is likely an artifact of the method of stimulus display.

Variable sampling window flip-flops for low-power high-speed VLSI

The authors describe novel flip-flops with improved robustness and reduced power consumption. The variable sampling window flip-flop (VSWFF) adjusts the width of the sampling window according to input data, providing robust data latching as well as shorter hold time. The flip-flop also reduces internal power consumption for high input switching activities. The clock swing-reduced variable sampling window flip-flop (CSR-VSWFF) reduces clock power consumption by allowing the use of a small swing clock. Unlike conventional reduced clock swing flip-flops, the proposed designs require no additional voltage higher than the supply voltage, eliminating the design overhead related to the generation and distribution of this voltage. Simulation results indicate that the proposed flip-flops provide uniform latency for a narrower sampling window and reduced power-delay product compared to conventional flip-flops. To evaluate the performance of the proposed flip-flops, test structures were designed and implemented in 0.35 /spl mu/m CMOS process technology. Experimental results indicate that the VSWFF consumes less power under conditions of maximum input switching activity. A synchronous counter designed using the CSR-VSWFF provides improved performance in terms of power consumption with no use of an extra voltage higher than the supply voltage.

Different patterns of excitation and inhibition of the small hand and forearm muscles from magnetic brain stimulation in humans.

The objective of the present study is to quantify the effects of voluntary muscle contraction of the small hand (abductor pollicis brevis, first dorsal interosseus (FDI)) and forearm muscles (extensor carpi radialis (ECR), extensor carpi ulnaris (ECU), flexor carpi radialis (FCR) and flexor carpi ulnaris (FCU)) on motor evoked potentials (MEPs). MEPs were recorded in 12 healthy subjects by a circular coil placed over the vertex at 1.2 times the resting motor threshold at different levels of the muscle contraction (0-100% of maximum voluntary contraction (MVC)). The effects of transcranial magnetic stimulation (TMS) on the onset latency, MEP area and silent period (SP) as a function of the %MVC were evaluated using a piecewise linear regression analysis. The MEP areas for the small hand muscles were almost completely saturated at 20% of MVC. In contrast, the MEP areas for radial muscles (ECR, FCR) had a dual increase at 40% of MVC while the ulnar muscles (ECU, FCU) had a dual increase at 20% of MVC. A uniform latency shift (1.5-3 ms reduction) was observed in all muscles with a changing point at 10% of MVC. The SPs were the longest for FDI and were not significantly influenced by MVC for any muscles. The excitatory and inhibitory effects of TMS on the MEPs differed for the small hand and forearm muscles and also between the ulnar and radial muscles. These results probably resulted from the different degrees of direct corticomotoneuronal inputs to each muscle and the inherent properties of the spinal motoneurons.

Gamma interferon can prevent herpes simplex virus type 1 reactivation from latency in sensory neurons.

We recently demonstrated that CD8(+) T cells could block herpes simplex virus type 1 (HSV-1) reactivation from latency in ex vivo trigeminal ganglion (TG) cultures without destroying the infected neurons. Here we establish that CD8(+) T-cell prevention of HSV-1 reactivation from latency is mediated at least in part by gamma interferon (IFN-gamma). We demonstrate that IFN-gamma was produced in ex vivo cultures of dissociated latently infected TG by CD8(+) T cells that were present in the TG at the time of excision. Depletion of CD8(+) T cells or neutralization of IFN-gamma significantly enhanced the rate of HSV-1 reactivation from latency in TG cultures. When TG cultures were treated with acyclovir for 4 days to insure uniform latency, supplementation with recombinant IFN-gamma blocked HSV-1 reactivation in 80% of cultures when endogenous CD8(+) T cells were present and significantly reduced and delayed HSV-1 reactivation when CD8(+) T cells or CD45(+) cells were depleted from the TG cultures. The effectiveness of recombinant IFN-gamma in blocking HSV-1 reactivation was lost when its addition to TG cultures was delayed by more than 24 h after acyclovir removal. We propose that when the intrinsic ability of neurons to inhibit HSV-1 gene expression is compromised, HSV-specific CD8(+) T cells are rapidly mobilized to produce IFN-gamma and perhaps other antiviral cytokines that block the viral replication cycle and maintain the viral genome in a latent state.

A highly-scalable, rapidly-reconfigurable, multicasting-capable, 100-Gb/s photonic switched interconnect based upon OTDM technology

We describe an ultrafast photonic switched interconnect based upon technologies developed for optical time division multiplexing (OTDM). The system uses a time-interleaved broadcast-and-select star architecture that is functionally equivalent to a crossbar switch. The interconnect offers full connectivity and low uniform latency among the input and output ports. The enabling technologies include ultrafast gated time slot tuners and all-optical demultiplexers. By utilizing these advanced optical technologies, it is possible to construct a highly scalable, rapidly reconfigurable, ultra-high-speed switch with performance beyond the capacity of current electronics. In the experimental demonstration, we constructed an interconnect with a peak bit rate of 100 Gh/s and the capability of connecting 16 OTDM ports. The system successfully demonstrated error-free operation of 100 Gb/s-multiplexing and demultiplexing in addition to rapid inter-channel switching capability on the order of the single channel bit period. The system also supports multicasting functions among many nodes. To scale the system to accommodate a large number of ports, we provide an analysis of the coherent crosstalk requirements through the network to show the potential to support hundreds of ports within practical constraints of the optical components. We believe that this system offers an approach to meet the demands of high bandwidth and fast switching capability required in current high-speed lightwave networks.

Determining the site of stimulation during magnetic stimulation of a peripheral nerve

Magnetic stimulation has not been routinely used for studies of perioheral nerve conduction primarily because of uncertainty about the location of the stimulation site. We performed several experiments to locate the site of nerve stimulation. Uniform latency shifts, similar to those that can be obtained during electrical stimulation, were observed when a magnetic coil was moved along the median nerve in the region of the elbow, thereby ensuring that the properties of the nerve and surrounding volume conductor were uniform. By evoking muscle responses both electrically and magnetically and matching their latencies, amplitudes and shapes, the site of stimulation was determined to be 3.0 ± 0.5 cm from the center of an 8-shaped coil toward the coil handle. When the polarity of the current was reversed by rotating the coil, the latency of the evoked response shifted by 0.65 ± 0.05 msec, which implies that the site of stimulation was displaced 4.1 ± 0.5 cm. Additional evidence of cathode- and anode-like behavior during magnetic stimulation comes from observations of preferential activation of motor responses over H-reflexes with stimulation of a distal site, and of preferential activation of H-reflexes over motor responses with stimulation of a proximal site. Analogous behavior is observed with electrical stimulation. These experiments were motivated by, and are qualitatively consistent with, a mathematical model of magnetic stimulation of an axon.

High susceptibility of FVB/N mice to the paralytic disease induced by ts1, a mutant of Moloney murine leukemia virus TB

The ts1 mutant of Moloney murine leukemia virus TB causes a degenerative neurologic and immunologic disease in susceptible strains of mice. This disease syndrome is characterized by development of spongiform encephalomyelopathy resulting in hindlimb paralysis, generalized bodywasting, and marked thymic atrophy associated with immune deficiency. The viral genetic determinants responsible for hindlimb paralysis in BALB/c and CFW/D mice have been localized to two point mutations in the env gene: one results in a Val-25-Ile substitution in the envelope precursor polyprotein gPr80 env and the other, in an Arg-430→Lys substitution in the gp70. In this report we present studies showing that FVB/N mice were highly susceptible to ts1 and exhibited the shortest and most uniform latency period of all the murine strains tested. In addition, we have found that, unlike in CFW/D and BALB/c mice, only the Val-25→lle substitution in the gPr80 env is required to induce hindlimb paralysis in FVB/N mice. Our studies show that there was enhanced replication of ts1 in all tissues of FVB/N mice and that the virus titer in the spinal cord was more than 10-fold higher in FVB/N than in BALB/c mice by 30 days postinoculation, when the clinical signs of paralysis became evident in FVB/N mice. Apparently, other host factors that do not require the Arg-430→Lys substitution allowed high levels of viral replication within the central nervous system of FVB/N mice. These results, together with the finding that 100% of FVB/N mice that were inoculated with ts1 at 5 days of age developed hindlimb paralysis at 30–60 days postinoculation, whereas only 33% of 5-day-old BALB/c mice developed hindlimb paralysis with a much longer latency period, suggest that subtle virus-host interactions determine the incidence, the latency period, and the severity of the disease caused by ts1.

Effects of spinal cord compression on repetitive impulse conduction of the ascending fibers of the dorsal column.

The effects of spinal cord compression on conduction of compound action potentials (CAPs) in the dorsal column fibers at various stimulus frequencies were analysed in pentobarbital anesthetised cats. The L 6 posterior root was given 1 to 500 Hz stimuli, and the CAPs were recorded from the L 2 posterior column. The 100th CAP was recorded at 100, 200 and 500 Hz. The L 4 cord segment was compressed by stepped increments (0.5 mm/5 min.) with a plastic plate (4 x 10 mm) until the responses at 1 Hz disappeared, then the compression was released. Before compression, each CAP at all frequencies showed almost uniform amplitudes and latencies. During compression, the CAPs were not altered significantly at any frequency until the spinal cord was compressed 2.5-4.0 mm. Further compression produced a progressive decrease in amplitude and increase in latency of the CAPs. At 500 Hz, however, the conduction block (amplitude loss, latency increase) was much severer than that produced at lower frequencies. After release of compression, the CAPs at all frequencies recovered progressively, and became almost stable after about 1 hour. The recovery rate of amplitude was only about 30% compared to the amplitude before compression at 500 Hz, and about 50% at 1 and 200 Hz. However, the recovery rate of amplitude at 100 Hz (65%) exceeded those at 1 and 200 Hz. To observe the difference in recovery rates among them, all the trains of CAPs were analyzed. Before compression, the amplitudes were almost uniform at stimulus frequencies as high as 500 Hz.(ABSTRACT TRUNCATED AT 250 WORDS)

Activation of fine articular afferent units by bradykinin

The responses of single fine articular afferent units to close intra-arterial injection of KCl and bradykinin were recorded from filaments of the saphenous nerve of the cat's right hindlimb. All units included in this study were sensitive to local mechanical probing of the medial and anteromedial aspects of the knee joint. The units were identified by conduction velocity as belonging either to group III (2.5–20 m/s, 17 units) or group IV (< 2.5 m/s, 23 units). Prior to bradykinin administration the responses of all units to passive limb movements were recorded in order to classify the units as belonging to one of the following 4 categories: activated by non-noxious movements; weakly activated by non-noxious movements; activated only by noxious movements; not activated by movements. Bolus injections of KCl were used to test the accessibility of the units via the blood vessels. Such injections elicited a rapid burst of impulses at short latencies of less than 1 s. If this discharge did not occur, no test with bradykinin was carried out. There was no difference in latency and time course between such discharges in group III and group IV units. With only 3 exceptions the 40 units excited by KCl were also activated by bradykinin which was applied in doses from 0.026 to 26 μg. Higher doses were not used. For most group III and IV units the minimal effective dose of bradykinin for a clear excitation was usually either 0.26 or 2.6 μg. In both groups of units the bradykinin-evoked discharge generally had a uniform latency and a time course with a total duration well under 1 min. In the course of repetitive injections at intervals of 3–5 min, the latency of the evoked discharge increased gradually and its magnitude became successively smaller. This tachyphylaxis was usually very pronounced, regardless of whether low or high doses of bradykinin were administered. No differences in the bradykinin sensitivity were found between units with very low to very high thresholds to local mechanical stimulation (tested with von Frey hairs) and between units belonging to the 4 different categories of response behavior to passive innocuous and noxious joint movements. These results indicate that the sensitivity to bradykinin is shared by all fine articular afferent units, regardless of their thresholds to local mechanical stimulation and joint movement and, hence, their functional role in signaling innocuous or noxious mechanical events at the knee joint. In contrast, fast group II articular units (conduction velocity > 20 m/s) are readily excited by KCl, but not by bradykinin administration.

The gravireceptor of Phycomyces. Its development following gravity exposure.

The gravitropism of a mature stage IV Phycomyces sporangiophore has a shorter and more uniform latency if the sporangiophore is exposed horizontally to gravity during its earlier development (stage II and stage III). This early exposure to an altered gravitational orientation causes the sporangiophore to develop a gravireceptor as it matures to stage IV and resumes elongation. A technique has been developed to observe the spatial relationship between the vacuole and the protoplasm of a living sporangiophore and to show the reorganization caused by this exposure to altered gravity. Possible gravireceptor mechanisms are discussed.

Scheduling Trees in Parallel/Pipelined Processing Environments

Scheduling task trees to be executed in parallel and/or pipelined processing systems are examined under individual situations. Processor structural requirements at task nodes are also included in the model of consideration. While simple techniques can serve as heuristics, counterexamples are constructed in some crucial cases. Simple optimal algorithms are presented in two important cases: 1) unistructure, multipipe, uniform latency, and flush time; and 2) vector loops. Finally, the complexity of the remaining cases is scrutinized with different structural parameter combinations.

Iron-dextran induction of distant tumours in mice.

AbstractThe systemic carcinogenicity of iron‐dextran has been investigated after single and repeated injections of Imferon in mice and rats. Systemic iron deposits induced by irondextran resulted in a stepwise development of reticulosarcomas. Such a histologic reaction, however, was not seen in relation to iron deposits following iron‐carbohydrate complex (Intrafer) administration. Distant tumours arising in the lymphoreticular structures showed a uniform latency in test and control mice, while tumours of other origin showed a prolonged latency as compared to such tumours in control animals. It is suggested, therefore, that iron‐dextran may act as a co‐carcinogenic factor in the case of lymphoreticular neoplasia, increasing host susceptibility to the oncogenic virus presumably present in both control and test animals. The non‐lymphoreticular distant and local tumours may stem from a different and direct carcinogenic action of iron since these tumours showed a prolonged latency and were seen after the administration of either iron‐dextran or iron‐carbohydrate complex.The results indicate an inconclusive dose‐response relationship, though no significant difference was found between tumour yield in the larger groups receiving 500 and 25 mg Fe…/kg body weight, the latter dose being within the therapeutic level on a comparative weight basis.