Search

FPGAs can be a promising accelerator used for MEC (Multi-access Edge Computing) which provides timing critical services for a number of terminals at the base stations near from edges. Although a high-end FPGA can support a fixed latency computation with a relatively small power consumption, they are expensive and the available acceleration circuits are limited into a size of single FPGA. FiC (Flow-in-Cloud) has been developed for building a virtual large FPGA from a number of middle-range economical FPGAs connected with high speed serial links. Although the current target of FiC is cloud computing, it is more suitable for the future MEC, because huge hardware resource can be supported with small cost. One of the problem to use such multi-FPGA systems for timing critical computation is network uncertainty. With a common packet switching, the computation speed is influenced with the network traffic. That is, the fixed latency computation which could be supported by a single FPGA is hard to be supported with multi-FPGA systems using common packet switching networks. In order to address this problem, we introduced STDM (Static Time Division Multiplexing) switch in the FiC system. Since the STDM always supports a constant communication latency, transfer time can be estimated beforehand. Through the implementation of the STDM switch on the FPGA board for FiC, it appeared that the utilization ratio of the LUTs for the STDM switch is smaller than 14%. The required number of slots is less than 16 even for a system with 256 nodes. We implemented the Conjugate Gradient method, which includes all-to-all communication, on 4x2 FiC system. It achieved 17.9 times performance improvement of Intel E5-2667 2.90GHz CPU with 6 cores.

We developed a power converter for small hydroelectric generation. It is an AC/DC/AC conversion system with an inverter for permanent magnetic generator control and an inverter for system interconnection contained in one package. Sensorless PM rotating machine speed control with MPPT was applied to generator control. For MPPT, A similar generator control system for solar power generation was applied and stability control was added while accounting for the fact that hydroelectric generation is subject to higher fluctuation in the power generation amount than solar power generation. In addition, a protective function was added to prevent the mechanical system from entering a critical speed range such as overspeed or mechanical resonance speed.

Sensor nodes used in Internet of Things (IoT) are required to work an extremely long time without replacing the battery. Natural renewable energy such as a solar battery is a hopeful candidate for such nodes. Here, a power model for operating an Silicon on Insulator (SOI) device with a solar battery including a large inner resistance is proposed, and applied to a micro-controller V850E-star and an accelerator CMA-SOTB2. Unlike the ideal case, the maximum operational frequency was achieved with reverse biasing by suppressing the leakage current which decreases the supply voltage. Under the room light with a large inner resistance, the strong reverse bias is effective, while a relatively weak reverse bias is advantageous under the bright light. The proposed model is appeared to be useful to estimate the appropriate body bias voltage both for V850E-star and CMA-SOTB2. In the V850E-star, the estimated operational frequencies were different from the real chip, while they were relatively matched when CMA-SOTB2 was used under the low illuminance.

The purpose of this paper is to present multiple motors connected in parallel are driven with stability. The parameters of each motor, namely stator current, rotor flux, and speed were estimated using adaptive rotor flux observers to achieve sensorless control. The validity and effective of the proposed method was demonstrated through simulations.

The purpose of this paper is to present multiple motors connected in parallel are driven with stability. The parameters of each motor, namely stator current, rotor flux, and speed were estimated using adaptive rotor flux observers to achieve sensorless control. The validity and effective of the proposed method was demonstrated through simulations.

Given an undirected hypergraph G/sub 0/ with two vertices s and t specified, our problem is to extract a maximal subgraph under a constraint. It comprises two terms, the upper bound of the area and the cut separating the subgraph being an s-t minimum cut. If G/sub 0/ is a model of a logic circuit, this is a faithful problem formulation in implementation of PPGA or MCM. Aiming to support such a design, we propose a representation of the s-t minimum cut of G/sub 0/ in terms of a directed acyclic graph G*. The key idea is to transform G/sub 0/ to a directed flow-graph with source s and sink t by a known technique and, after finding a maximum flow, to traverse the graph by a forward-search, which we introduce here.

This paper presents a formula to count the number of occurrence of interval parameters in the expression of an interval solution of a linear circuit equation when we use the Gaussian algorithm. The effect of the occurrence is shown to be related to the choice of a tree in the hybrid formulation of the circuit.

Performance requirements are soaring for embedded processors, whose demand in multimedia processing is rising now more than ever. Some DSP and media processors satisfy this by means of VLIW architecture. However, for embedded processors, less code, low power and small die are compulsory. These requirements make 4-way super-scalar embedded processor impractical. This embedded processor utilizes 4-way VLIW architecture characterized by: (1) parallel execution by VLIW. (2) generic CPU function in combination with media processing function for enhancing multimedia processing ability. (3) NOP instruction suppressing by packing flags for compatibility among different parallel levels. (4) two parallel execution mechanisms, ILP and SIMD.

The drying characteristics of municipal solid waste were experimentally examined. A mixture of pulp (water soluble pulp), plastic (polyvinyl chloride powder), protein (soybean protein powder) and water was used as the model refuse. The operational conditions were varied and the changes in weight and temperature with time were measured. The hard layer is formed on the drying side of the sample due to protein denaturation. With increasing CaO content, this hard layer becomes thinner and more cracks are generated. As a result, when the temperature of drying gas is relatively high, the drying rate is higher in the case with CaO than in the case without CaO. It is found that the uniform addition of CaO to the refuse and CO/sub 2/ in the drying gas enhance the drying rate in the RDF production.

The memory retention life and its repeatability was examined under conditions of high temperature storage, temperature cycling, bias operating, and high humidity, and after write/erase cycles for MNMoOS EAROMs having short memory retention life. It is shown that the high temperature storage and humidity tests are representative of all these tests, retention life during humidity tests for PED's is generally shorter than the unrecoverable failure life, and SiN overcoating extends retention life remarkably.