Abstract: A caching system is provided. The computing infrastructure runs off of a centralized storage, and data stored on the centralized store can also be retrieved from nearby machines that are part of the local infrastructure and have recently accessed the centralized store. Address-to-digest mappings are used to find an index of the desired data block. That digest is then used to hold where the data block is being cached. In some embodiments, the digest is hashed and the hash of the digest is used to determine where the data block is being cached. The data block is accessed from the cache using its cache, therefore different addresses may result in the retrieval of the same data block. For example, in a virtual machine environment, two different nodes may retrieve the same data block using different addresses.

Abstract: Systems, methods, and products for pull data transfer in a request-response model are provided herein. One aspect provides for generating output data utilizing at least one data generation station; and communicating via the at least one data generation station output data related to at least one data request received from at least one data requesting station responsive to at least one criterion, the at least one criterion comprising one of expiration of a time period or generation of a threshold amount of output data. Other embodiments and aspects are also described herein.

Abstract: A method and computer program product for assembling and deploying multi-platform flow based applications. An information processing flow that produces a result is assembled, the information processing flow includes components connected by data links, a component includes software code that describes at least one of an input constraint or an output constraint of the component, and at least two of the components are deployable on different computing platforms. The information processing flow is partitioned into sub-flows, such that for each sub-flow every component in the sub-flow is deployable on the same computing platform. The sub-flows are deployed on their respective computing platforms.

Abstract: A caching system is provided. The computing infrastructure runs off of a centralized storage, and data stored on the centralized store can also be retrieved from nearby machines that are part of the local infrastructure and have recently accessed the centralized store. Address-to-digest mappings are used to find an index of the desired data block. That digest is then used to hold where the data block is being cached. In some embodiments, the digest is hashed and the hash of the digest is used to determine where the data block is being cached. The data block is accessed from the cache using its cache, therefore different addresses may result in the retrieval of the same data block. For example, in a virtual machine environment, two different nodes may retrieve the same data block using different addresses.

Abstract: Systems, methods, and products for pull data transfer in a request-response model are provided herein. One aspect provides for generating output data utilizing at least one data generation station; and communicating via the at least one data generation station output data related to at least one data request received from at least one data requesting station responsive to at least one criterion, the at least one criterion comprising one of expiration of a time period or generation of a threshold amount of output data. Other embodiments and aspects are also described herein.

Abstract: A guide wire arrangement, a strip arrangement, a method of forming a guide wire arrangement, and a method of forming a strip arrangement are provided. The guide wire arrangement includes a strip; a sensor being disposed on a first portion of the strip; a chip being disposed next to the sensor on a second portion of the strip, wherein the second portion of the strip is next to the first portion of the strip; wherein the strip is folded at a folding point between the first portion of the strip and the second portion of the strip such that the first portion of the strip and the second portion of the strip form a stack of strip portions.

Abstract: An apparatus to facilitate design of a stream processing flow that satisfies an objective, wherein the flow includes at least three processing groups, wherein a first processing group includes a data source and an operator, a second processing group includes a data source and an operator and a third processing group includes a join operator at its input and another operator, wherein data inside each group is organized by channels and each channel is a sequence of data, wherein an operator producing a data channel does not generate new data for the channel until old data of the channel is received by all other operators in the same group, and wherein data that flows from the first and second groups to the third group is done asynchronously and is stored in a queue if not ready for processing by an operator of the third group.

Abstract: In a method for visualizing query results in stream processing systems, a visualization service receives a query from a client to visualize data in a stream processing application. The query is sent from the visualization service to a query-able operator of the stream processing application. At the query-able operator, an operation is performed using history data in the query-able operator to produce a first result that satisfies the query and the first result is sent to the visualization service. At the query-able operator, another operation is performed using new data received by the query-able operator to produce a second result that satisfies the query and the second result is sent to the visualization service. The first and second results are output from the visualization service to the client.

Abstract: In a method for visualizing query results in stream processing systems, a visualization service receives a query from a client to visualize data in a stream processing application. The query is sent from the visualization service to a query-able operator of the stream processing application. At the query-able operator, an operation is performed using history data in the query-able operator to produce a first result that satisfies the query and the first result is sent to the visualization service. At the query-able operator, another operation is performed using new data received by the query-able operator to produce a second result that satisfies the query and the second result is sent to the visualization service. The first and second results are output from the visualization service to the client.

Abstract: A method and computer program product for assembling and deploying multi-platform flow based applications. An information processing flow that produces a result is assembled, the information processing flow includes components connected by data links, a component includes software code that describes at least one of an input constraint or an output constraint of the component, and at least two of the components are deployable on different computing platforms. The information processing flow is partitioned into sub-flows, such that for each sub-flow every component in the sub-flow is deployable on the same computing platform. The sub-flows are deployed on their respective computing platforms.

Abstract: Methods and apparatus operating in a stream processing network perform load shedding and dynamic resource allocation so as to meet a pre-determined utility criterion. Load shedding is envisioned as an admission control problem encompassing source nodes admitting workflows into the stream processing network. A primal-dual approach is used to decompose the admission control and resource allocation problems. The admission control operates as a push-and-pull process with sources pushing workflows into the stream processing network and sinks pulling processed workflows from the network. A virtual queue is maintained at each node to account for both queue backlogs and credits from sinks. Nodes of the stream processing network maintain shadow prices for each of the workflows and share congestion information with neighbor nodes.

Abstract: Methods and apparatus operating in a stream processing network perform load shedding and dynamic resource allocation so as to meet a pre-determined utility criterion. Load shedding is envisioned as an admission control problem encompassing source nodes admitting workflows into the stream processing network. A primal-dual approach is used to decompose the admission control and resource allocation problems. The admission control operates as a push-and-pull process with sources pushing workflows into the stream processing network and sinks pulling processed workflows from the network. A virtual queue is maintained at each node to account for both queue backlogs and credits from sinks. Nodes of the stream processing network maintain shadow prices for each of the workflows and share congestion information with neighbor nodes.

Abstract: An apparatus, including a memory device for storing a program, and a processor in communication with the memory device, the processor operative with the program to facilitate design of a stream processing flow that satisfies an objective, wherein the stream processing flow includes at least three processing groups, wherein a first processing group includes a data source and an operator, a second processing group includes a data source and an operator and a third processing group includes a join operator at its input and another operator, wherein data inside each group is organized by channels and each channel is a sequence of data, wherein an operator producing a data channel does not generate new data for the channel until old data of the channel is received by all other operators in the same group, and wherein data that flows from the first and second groups to the third group is done asynchronously and is stored in a queue if not ready for processing by an operator of the third group, and deploy the stream

Abstract: A microfluidic filtration unit for trapping particles of a predetermined nominal size present in a fluid is provided. The unit comprises a fluid chamber connected to an inlet for introducing the fluid to be filtered and an outlet for discharging filtered fluid, a filtration barrier arranged within the fluid chamber, said filtration barrier comprising a plurality of pillars arranged substantially perpendicular to the path of fluid flow when fluid is introduced into the fluid chamber, said pillars being aligned to form at least one row extending across said path of fluid flow, wherein each of said at least one row of pillars in the filtration barrier comprises at least one fine filtration section comprising a group of pillars that are spaced apart to prevent particles to be filtered from the fluid from moving between adjacent pillars, and at least one coarse filtration section comprising a group of pillars that are spaced apart to permit the movement of particles between adjacent pillars.

Abstract: Methods and apparatus operating in a stream processing network perform load shedding and dynamic resource allocation so as to meet a pre-determined utility criterion. Load shedding is envisioned as an admission control problem encompassing source nodes admitting workflows into the stream processing network. A primal-dual approach is used to decompose the admission control and resource allocation problems. The admission control operates as a push-and-pull process with sources pushing workflows into the stream processing network and sinks pulling processed workflows from the network. A virtual queue is maintained at each node to account for both queue backlogs and credits from sinks. Nodes of the stream processing network maintain shadow prices for each of the workflows and share congestion information with neighbor nodes.

Abstract: Methods and apparatus operating in a stream processing network perform load shedding and dynamic resource allocation so as to meet a pre-determined utility criterion. Load shedding is envisioned as an admission control problem encompassing source nodes admitting workflows into the stream processing network. A primal-dual approach is used to decompose the admission control and resource allocation problems. The admission control operates as a push-and-pull process with sources pushing workflows into the stream processing network and sinks pulling processed workflows from the network. A virtual queue is maintained at each node to account for both queue backlogs and credits from sinks. Nodes of the stream processing network maintain shadow prices for each of the workflows and share congestion information with neighbor nodes.

Abstract: A silicon condenser microphone is described. The silicon condenser microphone of the present invention comprises a perforated backplate comprising a portion of a single crystal silicon substrate, a support structure formed on the single crystal silicon substrate, and a floating silicon diaphragm supported at its edge by the support structure and lying parallel to the perforated backplate and separated from the perforated backplate by an air gap.

Abstract: The invention relates to a method of making an integrated circuit inductor that comprises a silicon substrate and an oxide layer on the silicon substrate. In one aspect, the method comprises depositing an inductive loop on the oxide layer, and making a plurality of apertures in the oxide layer beneath the inductive loop. The method also comprises providing a plurality of bridges adjacent the apertures and provided by portions of the oxide layer between an inner region within the inductive loop and an outer region of the oxide layer without the inductive loop, the inductive loop being supported on the bridges. The method comprises forming a trench in the silicon substrate beneath the bridges, to provide an air gap between the inductive loop and the silicon substrate.

Abstract: When using hot alkaline etchants such as KOH, the wafer front side, where various devices and/or circuits are located, must be isolated from any contact with the etchant. This has been achieved by using two chambers that are separated from each other by the wafer that is to be etched. Etching solution in one chamber is in contact with the wafer's back surface while deionized water in the other chamber contacts the front surface. The relative liquid pressures in the chambers is arranged to be slightly higher in the chamber of the front surface so that leakage of etchant through a pin hole from back surface to front surface does not occur. As a further precaution, a monitor to detect the etchant is located in the DI water so that, if need be, etching can be terminated before irreparable damage is done.

Abstract: A silicon condenser microphone is described. The silicon condenser microphone of the present invention comprises a perforated backplate comprising a portion of a single crystal silicon substrate, a support structure formed on the single crystal silicon substrate, and a floating silicon diaphragm supported at its edge by the support structure and lying parallel to the perforated backplate and separated from the perforated backplate by an air gap.