Patents Examined by Agni Mohamed
-
Patent number: 5862365Abstract: An integrated circuit for providing a pin-for-pin replacement of a field programmable gate array (FPGA). The integrated circuit includes an emulation circuit for mimicking the programmable stage (e.g., initialization, configuration and start-up states) of the FPGA. The integrated circuit is designed to be transparent to the user/customer, thereby eliminating the need for a costly redesign of a user's circuit board.Type: GrantFiled: September 17, 1996Date of Patent: January 19, 1999Assignee: Lucent Technologies Inc.Inventors: Ronald T. Modo, Gary P. Powell, Hollis G. Robertson, William H. Smith III
-
Patent number: 5548735Abstract: A system and method for asynchronously managing the issuance of program I/O store instructions from a high speed central processor to a multiplicity of relatively lower speed I/O adapter devices. An interface between the central processor and the I/O adapter devices includes a program I/O store queue, a state machine, and a token pool related in count to the concurrent processing capabilities of I/O controllers. The interface queue includes information for uniquely identifying program I/O store instructions by adapter device destination and user application program to manage error recovery. As preferably implemented, the interface system and method also distinctly manages program I/O instructions requiring synchronous execution, such as program I/O load instructions.Type: GrantFiled: September 15, 1993Date of Patent: August 20, 1996Assignee: International Business Machines CorporationInventors: Wen-Tzer T. Chen, Steven M. Thurber, Gary Y. Tsao
-
Patent number: 5515525Abstract: A memory translation mechanism and method executing in a second system to perform first system memory operations for first system executive and user tasks executing on the second system which includes a second system memory organized as a plurality of memory segments, wherein first memory segments are designated to correspond to system memory areas and second memory segments are designated to correspond to user memory areas, and wherein each memory segment corresponds to a combination of a type of first system task and a type of a first system memory area. An interpreter maps by reading an identification of the type of the task corresponding to the first system virtual address from the task type memory and the area type value from the first system virtual address and determining a memory segment corresponding to the type of the first system task and the type of first system area referenced by the first system virtual address.Type: GrantFiled: September 28, 1993Date of Patent: May 7, 1996Assignee: Bull HN Information Systems Inc.Inventors: Marek Grynberg, Dennis R. Flynn, Thomas S. Hirsch, Mary E. Tovell, William E. Woods
-
Patent number: 5495594Abstract: By monitoring various combinations of control signals generated by a microprocessor in a computer system in the first operational cycles after it is reset, a peripheral circuit sets itself to respond appropriately to control signals from the microprocessor according to any of several different protocols. For example, an instruction from the microprocessor to write to or read from the peripheral circuit is implemented over two control lines with one of several possible protocols. The circuit determines which protocol is being used each time the system is initialized and thereafter knows when a read or write operation is being performed. Another example is the different wait or acknowledge protocols that various microprocessors use.Type: GrantFiled: May 24, 1994Date of Patent: February 27, 1996Assignee: Zilog, Inc.Inventors: Craig A. MacKenna, Monte J. Dalrymple
-
Patent number: 5469553Abstract: For a computer system or a subsystem thereof having electrical components, a method and apparatus for a collection of event driven software state machine of the type where each state machine is separately operable at differing levels of power consumption, and where the transitions from state to state are as a direct result of input events. Each of the state machines is programmatically biased to operate the state machine at a lowest possible power, and state machines processing event of a higher priority do so at the expense of state machines processing events having a lower priority.Type: GrantFiled: December 16, 1994Date of Patent: November 21, 1995Assignee: Quantum CorporationInventor: Paul R. Patrick
-
Patent number: 5414865Abstract: Each self-programming module in a network uses its input signals to obtain a sequence of received signals. Each received signal is used to obtain a control data item having a first or a second value. If the first value, the received signal is also used to obtain a generated data item. If the second value, a learning response can occur, storing a limited length sequence of previous generated data items so that each can be accessed with data items having the same value as the received signal for which it was obtained. A generated data item can be obtained by accessing a stored previous generated data item. If none has been stored for the current received signal or in case of repetition of received signals or a long gap between learning responses, a random number can be obtained as the generated data item. When the control data item has the second value, the module can provide an output signal, and the output signal from one module can be an input signal to another.Type: GrantFiled: June 27, 1991Date of Patent: May 9, 1995Inventor: James T. Beran
-
Patent number: 5197131Abstract: In an information processing system having an instruction buffer, an instruction buffer is controlled to primarily increase the instruction hit ratio of a sequence of instructions including a procedure-call instruction. In the first configuration, there is provided a mechanism which subdivides the instruction buffer into a plurality of instruction buffer banks so as to switch the instruction buffer bank to a current use in association with a dynamic procedure call, thereby improving the instruction hit ratio in the procedure call and in the return operation. In the second configuration, there are provided instruction words to subdivide and to control the instruction buffer such that the user can specify a method of controlling the instruction buffer. An instruction loop is captured efficiently and an arbitrary instruction sequence of a program is stored as a resident routine in the buffer so as to increase the instruction hit ratio.Type: GrantFiled: February 8, 1989Date of Patent: March 23, 1993Assignee: Hitachi, Ltd.Inventors: Noriyasu Mori, Hiroshi Tomita
-
Patent number: 4896256Abstract: A plurality of data processor components are distributed along a plurality of intracommunication bus systems, each intracommunication bus system having an address bus and each being assigned a unique set of addresses, with at least one of the components associated with each of the intracommunication bus systems including an address signal generator for generating address signals over the address bus of the associated intracommunication bus system. An intercommunication bus system is utilized in combination with a plurality of link interface units, with each link interface unit connected between the intercommunication bus system and a corresponding one of the intracommunicaton bus systems, for carrying out communication of information over the intercommunication bus system between first and second of the intracommunication bus systems, in response to the address signals on the address bus of the corresponding one of the intercommunication bus systems. A related method is also provided.Type: GrantFiled: May 14, 1986Date of Patent: January 23, 1990Assignee: Kabushiki Kaisha ToshibaInventor: Barry R. Roberts