Actel Corporation, Inc.

Microsemi Soc Corporation

Gatefield Corporation

Volker Hecht

Jonathan W Greene

John L McCollum

Alan B Reynolds

Andrew W Reynolds

Theodore M Speers

Robert M Salter, Iii

Gregory William Bakker

Jack Zezhong Peng

Fethi Dhaoui

Kyung Joon Han

Vidyadhara Bellippady

Fei Li

Robert U Broze

Marcel Derevlean

Dirk Kannemacher

Zhigang Wang

Santosh Yachareni

Fengliang Xue

Victor Nguyen

Timothy Saxe

Kai Zhu

Victor Levchenko

Pavan Singaraju

Robert Ullrich Broze

Zhezhong Peng

A single-event-upset (SEU) stabilized memory cell includes a latch portion including a cross-coupled latch, and at least one cross coupling circuit path in the latch portion including a first series-connected pair of vertical resistors.

SEU stabilized memory cells

A 4-input lookup table module including eight first-rank 2-input multiplexers, four second-rank multiplexers, two third-rank multiplexers, and one fourth-rank multiplexer, the first-rank through fourth-rank multiplexers forming a tree structure. A select input of the fourth-rank multiplexer is coupled to a first input node. Select inputs of the third-rank multiplexers are coupled to a second input node. Select inputs of a first and a second adjacent ones of the second rank 2-input multiplexers are electrically isolated from select inputs of a third and a fourth adjacent ones of the second rank 2-input multiplexers. Select inputs of a first through a fourth adjacent ones of the first rank 2-input multiplexers are electrically isolated from select inputs of a fifth through an eighth adjacent ones of the first rank 2-input multiplexers.

Efficient lookup table modules for user-programmable integrated circuits

A method for preventing over-programming of resistive random access (ReRAM) based memory cells in a ReRAM memory array includes applying a programming voltage in a programming circuit path including a ReRAM memory cell to be programmed, sensing programming current drawn by the ReRAM cell while the programming voltage is applied across the memory cell, and decreasing the programming current as a function of a rise in programming current.

Circuits and methods for preventing over-programming of ReRAM-based memory cells

A ReRAM cell array has having at least one row and one column includes first and second complementary bit lines for each row, a word line, a p-word line, and an n-word line for each column. A ReRAM cell at each row and column of the array includes a first ReRAM device, its first end connected to the first complementary bit line of its row, a p-channel transistor, its source connected to a second end of the first ReRAM device, its drain connected to a switch node, its gate connected to the p-channel word line of its column, a second ReRAM device, its first end connected to the second complementary bit line of its row, an n-channel transistor, its source connected to a second end of the second ReRAM device, its drain connected to the switch node, its gate connected to the n-channel word line of its column.

Three-transistor resistive random access memory cells

A pair of adjacent ReRAM cells in an array includes a first bit line for a row of the array, a second bit line for the row of the array, a p-channel word line associated with two adjacent columns in the array, and an n-channel word line associated with the two adjacent columns. A pair of ReRAM cells in the adjacent columns in the row each includes a switch node, a first ReRAM device connected between the first bit line and the source of a p-channel transistor. The drain of the p-channel transistor is connected to the switch node, and its gate is connected to the p-channel word line. A second ReRAM device is connected between the second bit line and the source of an n-channel transistor. The drain of the n-channel transistor is connected to the switch node, and its gate is connected to the n-channel word line.

A fast lookahead carry adder includes adder logic and lookahead carry-path logic coupled to the adder logic. The carry path logic has a main carry path, a carry entrance path and a carry exit path, the carry entrance path separate from the carry exit path.

Fast carry lookahead circuits

An integrated programmable logic circuit having a read/write probe includes a plurality of programmable logic circuits having internal circuit nodes and a plurality of flip flops, each having an asynchronous data input line, an asynchronous load line, and a data output connected to an internal circuit node, a probe-data line, an address circuit for selecting one of the internal circuit nodes, a read-probe enable line for selectively coupling the selected one of the internal circuit nodes to the probe-data line, a data input path to the asynchronous data input line of each flip flop, a write-probe data input path to the asynchronous data input line of each flip flop, a write-probe enable line, and selection circuitry, responsive to the address circuit and the write-probe enable line, to couple one of the data input path and the write-probe data input path to the asynchronous data input of a selected flip flop.

On-chip probe circuit for detecting faults in an FPGA

An integrated circuit includes a clock input, a first output, and a second output. A programmable pulse generator has a programmable pulse counter coupled to the clock input at least one control input for receiving count information. A fixed delay element is coupled to the programmable pulse counter. A programmable delay element is coupled to the programmable pulse counter and has at least one control input for receiving delay information. A first multiplexer is coupled to the fixed delay element, the programmable delay element and to the first output. A second multiplexer is coupled to the programmable delay element, the output of the fixed delay element and the second output.

A random-access memory block for a field programmable gate array includes a random-access memory array having address inputs, a data input, a data output and including a plurality of storage locations. At least two programmably invertible enable inputs are provided. Hardwired decoding logic is coupled to the at least two programmably invertible enable inputs to selectively enable the random-access memory array. A gate is coupled to the output of the random-access memory array and is configured to pass the output of the random-access memory array only if the random-access memory is enabled for a read operation, and otherwise generate a preselected logic state.

RAM block designed for efficient ganging

A method for reducing delay in an integrated circuit by compensating for differences in rise and fall delay times comprises creating a timing graph; computing minimum delay tuples for nodes in the timing graph; if there is not at least one feasible delay tuple, determining a longest path and computing minimum delay tuples for the longest path; changing polarities on the longest path to reduce delays; updating the timing graph by transferring new polarity and delay values; performing timing analysis to determine a new longest path if the new longest path is shorter than the prior longest path, accepting a resulting polarity selection and computing minimum delay tuples for the longest path; if the new longest path is not shorter than the prior longest path, accepting a resulting polarity selection and implementing changes in a user-program bitstream.

Architecture and method for compensating for disparate signal rise and fall times by using polarity selection to improve timing and power in an integrated circuit

A flip-flop for use in a field programmable gate array integrated circuit device is disclosed. The flip-flop comprises a data output terminal coupled to a first programmable routing element, a data input terminal coupled to a second programmable routing element, and a clock input terminal, wherein a signal appearing at the data output terminal in response to a signal applied to the clock input terminal has the opposite logical polarity with respect to the corresponding logical signal applied to the data input terminal.

Inverting flip-flop for use in field programmable gate arrays

An FPGA architecture includes multiplexers having non-volatile switches having control gates coupled to word lines W, each word line associated with a row, the switches connecting to wiring tracks through buffers having a controllable ground connection NGND, at least some of the switches being a tie-off switch coupleable to one of a plurality of bitlines B, each bitline associated with column.

Circuits and methods for testing FPGA routing switches

A programmable logic device (PLD) with a plurality of programmable regions is disclosed. Some of the programmable regions have switch power or ground supplies to allow them to be put into a low-power state in one or more low-power modes. At least one of the programmable regions always remains on during the low-power modes to enable the user to design custom PLD power management logic that may be placed in the always-on programmable region.

Growing community of inventors

Barsinghausen, Germany

Volker Hecht