Abstract: An approach is described for a method, product, and apparatus for a machine learning process using weight sharing within a systolic array having reduced memory bandwidth. According to some embodiments, this approach includes providing a systolic array that includes processing elements which each have some number of storage elements for storing weights. For example, the weights can be reused for different data sets by identifying/capturing a current state of the storage elements, generating a plan to transition to a target state of those storage elements, and application of the transition plan such that weights that are already stored in those storage elements can be reused and/or relocate. This lowers the bandwidth requirements for weight memory by allowing weights that have previously been read into the systolic array to be reused.

Abstract: An approach includes receiving a machine learning processing job, executing the machine learning processing job using parallel processing of multiple output pixels each cycle by walking data across processing elements with broadcast weights within regions and executing parallel multiplication operations, and generating an output indicating whether the machine learning processing job was successful or failed. In some embodiments, a schedule of actions is generated for respective machine learning processing jobs. The schedule of actions may include any of a plurality of shift operations in a many to many arrangement or a one to many arrangement, shifting data across region boundaries, fetching data and weights from a memory and distribution thereof to a plurality of regions (e.g., weights are distributed to respective weight memories which subsequently broadcasts those weights in a specified order based on a schedule of actions, and where data is distributed to respective processing elements).

Abstract: An approach includes a method, product, and apparatus for dynamically removing sparse data on a pixel by pixel basis. In some embodiments, a machine learning processing job is received. The machine learning processing job is then executed on a pixel by pixel basis by selecting non-zero data values for input into a systolic array, wherein sparse data is not selected for input into the systolic array. Subsequently, a message is generated that provides an indication of whether the execution completed successfully. In some embodiments, the machine learning processing job comprises at least a plurality of multiply and accumulate operations. In some embodiments, at least one data value equal to zero for the machine learning processing job is not input into a systolic array. In some embodiments, a plurality of weights are input into a plurality of columns for each cycle.

Abstract: An approach is described for a method, product, and apparatus for a machine learning process using dynamic rearrangement of sparse data and corresponding weights. This approach includes a method, product, and apparatus for dynamically rearranging input data to move sparse data to a location such that computations on the sparse data might be avoided when executing a machine learning processing job. For example, sparse data within each row of the input matrix can be moved to the end of each corresponding row. When the input data is folded to fit the array, that sparse data might be at least partially contained within a fold that comprises only sparse data and possibly filler data. In such an event, computations on the fold are unnecessary and are avoided. In some embodiments, the approach includes dynamically rearranging a weight matrix to maintain a correspondence between the input data and the weights.

Abstract: An approach includes identification of a machine learning model for processing and generating an ordered set of weights with varying precisions and metadata that specifies where those values can be found in order to allow the identification of weights needed during processing. In a first embodiment, the variable precision weights are separated into different memory segments where each segment has weights of only a single precision. In a second embodiment, the variable precision weights are provided in a memory where weights of different precisions are intermingled, and those weights are identified using a sequence of pairs of data representing a number of weights with the same precision and the precision of those weights. In some embodiments, both the first and second embodiments are combined, where some segments contain weights with only a single precision and at least one segment stores weights with different precisions within a respective segment.

Abstract: Technologies are provided for creating and using template constraint expressions in constraint-based systems. Template constraint expressions can be created that can be used to define multiple usages of a same constraint rule in a configuration model. Using the template constraint expression, the constraint rule can be translated once and used multiple times as different instances of the rule are activated. Updates to the rule can be made to the template constraint expression and applied to all of the related instances. Constraint expressions can be created based on the template constraint expression. Multiple object instances in a configuration model can be identified that satisfy matching criteria of the template constraint expression. Variables of the matching object instances can be mapped to variable placeholders in the template constraint expression to create constraint expressions. A constraint solver can be used to evaluate the constraint expressions.

Abstract: Technologies are provided for creating and using template constraint expressions in constraint-based systems. Template constraint expressions can be created that can be used to define multiple usages of a same constraint rule in a configuration model. Using the template constraint expression, the constraint rule can be translated once and used multiple times as different instances of the rule are activated. Updates to the rule can be made to the template constraint expression and applied to all of the related instances. Constraint expressions can be created based on the template constraint expression. Multiple object instances in a configuration model can be identified that satisfy matching criteria of the template constraint expression. Variables of the matching object instances can be mapped to variable placeholders in the template constraint expression to create constraint expressions. A constraint solver can be used to evaluate the constraint expressions.

Abstract: The disclosure generally describes computer-implemented methods, software, and systems for using mock application data to validate applications. A particular application to be validated is identified, wherein the particular application is generated via a user interface for generating new applications, and wherein the particular application is consistent with a data model used during the generation of the particular application. A set of stored mock application data associated with the particular application is identified. At least a portion of the identified set of mock application data is incorporated into a runtime version of the particular application. The particular application is executed using the incorporated set of mock application data in a validation environment.

Abstract: A method includes transforming, by a report connector according to at least one input mapping rule of the report connector, first input data received by a first application to create second input data for a specific report of a plurality of existing reports of a second application. The method also includes providing, by the report connector, the second input data to a report application programming interface (API) capable of providing an interface to the report of the second application. The method also includes receiving, by the report connector, first output data from the report API, the first output data based on the output data from the specific report of the second application. The method also includes transforming, by the report connector according to at least one output mapping rule of the report connector, the first output data to create second output data for the first application.

Abstract: A method includes determining, by a conversion tool, at least one specific data associated with an interface of a service of a first application, the determining comprising retrieving metadata associated with the service of the first application, the metadata describing the specific data; and automatically generating, by the conversion tool, a data context for use in requesting the service from the first application, the data context comprising at least one data field for storing the at least one specific data associated with the interface of the service.

Abstract: Systems and methods are provided for enabling communication between a composite system providing additional functionality not contained in existing legacy systems and other existing systems using different commands, variables, protocols, methods, or instructions, when data may be located on more than one system. In an embodiment, multiple software layers are used to independently manage different aspects of an application. A business logic layer may be used in an embodiment to facilitate reading/writing operations on data that may be stored locally and/or on external systems using different commands, variables, protocols, methods, or instructions. A backend abstraction layer may be used in an embodiment in conjunction with the business logic layer to facilitate communication with the external systems. A user interface layer may be used in an embodiment to manage a user interface, a portal layer to manage a user context, and a process logic layer to manage a workflow.

Abstract: The semantic message maintenance system provides a tool for a system with a backend and set of intermediate systems to maintain semantics of a message such that the end user can receive the information provided by the backend system in a manner that is comprehendible to the end user who has no specific knowledge of the functioning of the backend system. A message mapping system uses a series of mapping definitions and standardized identification fields. Both variable data textual data can be preserved.

Abstract: Systems and methods are provided for enabling communication between a composite system providing additional functionality not contained in existing legacy systems and other existing systems using different commands, variables, protocols, methods, or instructions, when data may be located on more than one system. In an embodiment, multiple software layers are used to independently manage different aspects of an application. A business logic layer may be used in an embodiment to facilitate reading/writing operations on data that may be stored locally and/or on external systems using different commands, variables, protocols, methods, or instructions. A backend abstraction layer may be used in an embodiment in conjunction with the business logic layer to facilitate communication with the external systems. A user interface layer may be used in an embodiment to manage a user interface, a portal layer to manage a user context, and a process logic layer to manage a workflow.

Abstract: A method includes transforming, by a report connector according to at least one input mapping rule of the report connector, first input data received by a first application to create second input data for a specific report of a plurality of existing reports of a second application. The method also includes providing, by the report connector, the second input data to a report application programming interface (API) capable of providing an interface to the report of the second application. The method also includes receiving, by the report connector, first output data from the report API, the first output data based on the output data from the specific report of the second application. The method also includes transforming, by the report connector according to at least one output mapping rule of the report connector, the first output data to create second output data for the first application.

Abstract: A method includes determining, by a conversion tool, at least one specific data associated with an interface of a service of a first application, the determining comprising retrieving metadata associated with the service of the first application, the metadata describing the specific data; and automatically generating, by the conversion tool, a data context for use in requesting the service from the first application, the data context comprising at least one data field for storing the at least one specific data associated with the interface of the service.

Abstract: A method for creating a custom work list includes executing a query based on search criteria. Results of the query comprising at least one of a business object and a task conforming to the search criteria. The search criteria and the returned results are combined to create said custom work list.

Abstract: A flatware dispensing unit for dispensing flatware is disclosed including a primary housing having at least one compartment for receiving flatware therein with the compartment having an elongated slot for exposing at least a portion of the flatware. In a first embodiment, a cartridge having a complimentary elongated slot substantially aligned with the elongated slot in the compartment when the cartridge is positioned within the compartment. The flatware cartridge includes a widened portion which when inserted into the compartment aligns with a widened portion of the elongated slot of the compartment. In doing so, the cartridge may be easily dropped into the compartment such that the flatware need not be individually handled by the person stocking the dispenser unit and limits the number of utensils which can be removed from the dispenser at one time.

Abstract: The steel having improved deep-drawing properties is characterized in that it contains carbon in a proportion of less than 0.015%, manganese in a proportion of from 0.15 to 0.25%, sulfur in a proportion of less than 0.012% and aluminum in a proportion of less than 0.04%.This steel is intended for the production of thin sheet metal intended for deep-drawing, in accordance with a process comprising, in particular, the following operations:production, in a converter, of a steel having the above composition;hot-rolling entirely in the austenitic region;winding at a temperature higher than 650.degree. C.;continuous annealing, after cold-rolling, at a temperature below 700.degree. C.