Lattice Semiconductor Corporation

Chan-Chi Jason Cheng

Ju Shen

Om P Agrawal

Satwant Singh

Ting Yew

Qin Wei

San-Ta Kow

Albert L Chan

Ann Wu

Gregg R Josephson

Roy D Darling

Manish Garg

Brad Sharpe-Geisler

Cyrus Y Tsui

Jack Tim Wong

Chi Minh Nguyen

Benny Ma

Sam Tsai

Bradley Felton

Timothy Scott Swensen

Paul R Bonwick

Andrew Armitage

Giai Trinh

Yi Chiang

In one embodiment, a programmable logic device (PLD) with configuration memory includes at least one configuration memory cell and soft error detection (SED) logic for checking for errors in data stored by the configuration memory. The SED logic calculates a present data value for the configuration memory for comparison with a pre-calculated data value. A fuse within the PLD is configurable in a first logic state to enable the SED logic to read from the configuration memory cell in calculating the present data value and configurable in a second logic state to prevent the SED logic from reading from the configuration memory cell in calculating the present data value. The SED logic may be tested for correct operation by writing data representing a soft error into the configuration memory cell and enabling the SED logic to read from the configuration memory cell in calculating the present data value.

Testing of soft error detection logic for programmable logic devices

A programmable logic device, in accordance with one embodiment, includes a plurality of configuration memory cells, wherein at least one configuration memory cell is adapted to function as random access memory. Read/write circuitry writes to and reads from a corresponding first port of the configuration memory cells, including reading from the at least one configuration memory cell adapted to function as random access memory. Soft error detection logic checks for an error in data values stored by the plurality of configuration memory cells, including the at least one configuration memory cell adapted to function as random access memory. The soft error detection logic, for example, may thus be tested by changing a data value stored in the at least one configuration memory cell.

Soft error detection logic testing systems and methods

In one embodiment, a method of converting an uncompressed bitstream into a compressed bitstream for a programmable logic device (PLD) is disclosed. The method includes embedding a first data frame from the uncompressed bitstream into the compressed bitstream, wherein the first data frame comprises a first data set; embedding a first instruction into the compressed bitstream to load the first data frame into a first row of configuration memory of the PLD at an address associated with the first data frame; identifying a second data frame in the uncompressed bitstream, wherein the second data frame comprises the first data set; and embedding a second instruction into the compressed bitstream to load the first data frame into a second row of the configuration memory at an address associated with the second data frame.

Compression and decompression of configuration data using repeated data frames

Various techniques are provided to compress and decompress configuration data for use with programmable logic devices (PLDs). In one example, a method includes embedding a first data frame comprising a data set from an uncompressed bitstream into a compressed bitstream. The method also includes embedding a first instruction to instruct a PLD to load the first data frame into a data shift register, embedding a second instruction to instruct the PLD to load a first address associated with the first data frame into an address shift register, and embedding a third instruction to instruct the PLD to load the first data frame from the data shift register into a first row of a configuration memory corresponding to the first address. The method further includes identifying a second data frame comprising the data set in the uncompressed bitstream, and embedding fourth and fifth instructions in place of the second data frame.

In one embodiment, a programmable logic device includes a plurality of configuration memory cells and at least one spare configuration memory cell adapted to store configuration data for a memory cell identified within the plurality of configuration memory cells (e.g., identified as a defective memory cell). An address shift register within the device is adapted to provide programming signals to the plurality of configuration memory cells via wordlines. A data shift register within the device is adapted to provide configuration data to the plurality of configuration memory cells via bitlines. The data shift register is further adapted to provide configuration data from the spare configuration memory cell to the identified configuration memory cell.

Redundant configuration memory systems and methods

A programmable logic device, in accordance with an embodiment of the present invention, includes a plurality of multiplexers, having fuse input terminals and input signal terminals, and a plurality of associated fuses providing fuse signals to the fuse input terminals to control selection of the input signal terminals. The fuses in a first state select a first input signal terminal of the input signal terminals, with a first multiplexer from the plurality of multiplexers receiving a first logic level signal at the first input signal terminal and providing the first logic level signal to the first input signal terminal of a first set of the plurality of multiplexers. The fuses associated with the first set are adapted to be programmed before the fuses associated with the first multiplexer.

Multiplexer initialization systems and methods

Systems and methods are disclosed directed to techniques with respect to defective configuration memory cells. For example, in accordance with an embodiment of the present invention, a programmable logic device includes a plurality of configuration memory cells; and at least one spare memory cell, wherein the at least one spare memory cell is adapted to store configuration data to provide to at least one defective configuration memory cell.

Systems and methods are disclosed herein to provide high fan-out signal routing. For example, in accordance with an embodiment of the present invention, a programmable logic device includes a plurality of logic blocks; an interconnect structure adapted to route signals among the logic blocks; and a secondary routing network adapted to route signals among the logic blocks. The secondary routing network may include a plurality of horizontal splines adapted to route signals within the programmable logic device; a plurality of vertical spline taps adapted to route signals within the programmable logic device; a plurality of common interface blocks adapted to route signals between the horizontal splines and the vertical spline taps; and a plurality of horizontal secondary branches adapted to route signals from the vertical spline taps to the logic blocks.

High fan-out signal routing systems and methods

A programmable logic block within a programmable logic device includes at least two interconnected slices, each of the interconnect slices including at least two interconnected lookup tables. Each interconnected lookup table is adapted to receive input signals from a routing structure and to provide a LUT output signal. At least one of the slices includes a register adapted to register the LUT output signal of a lookup table and at least another of the slices includes fewer such registers than lookup tables.

Programmable logic device with enhanced logic block architecture

In one embodiment of the invention, a programmable logic block within a programmable logic device includes: a plurality of lookup tables, each lookup table providing a combinatorial output signal; and a plurality of registers, each register being adapted to register a selected one of the combinatorial output signals, wherein the number of registers is less than the number of lookup tables.

A memory provides a programmable write port data width and an independently programmable read port data width. The independence between the programmable write port data width and the programmable read port data width is achieved without the use of a third clock domain.

FIFO memory with programmable data port widths

A programmable device includes a plurality of programmable blocks each associated with a distributed memory block. The programmable blocks may be configured as logic or memory. The addressing circuitry for each distributed memory block may be shared with its associated programmable block or may be separate.

CPLD with multi-function blocks and distributed memory

Cascadable logic block architectures are disclosed for programmable logic devices, such as for high density and high performance complex programmable logic devices. The logic block architectures provide, for example, clusters or groups of logic blocks that may have cascadable inputs and/or product terms to provide flexible logic width and/or depth capability. The logic block architecture may, for example, be implemented in conjunction with a multi-stage interconnect architecture to provide array fuse density and/or interconnect fuse density savings compared to conventional architectures.

Cascaded logic block architecture for complex programmable logic devices

An integrated circuit includes non-volatile and volatile memory, with the volatile memory controlling the integrated circuit's functionality. Various techniques are disclosed for programming the different types of memory through one or more data ports to provide in-system programmability and dynamic reconfigurability. External configuration devices are not required if the data from the non-volatile memory is transferred directly to the volatile memory.

Non-volatile and reconfigurable programmable logic devices

A structure and method for in-system programming of a programmable logic device are provided. The in-system programming structure provides one dedicated pin for in-system programming function, additional in-system programming pins are multiplexed with programmable input/output pins used in functional operations. When an enable signal is received at the dedicated pin, the multiplexed pins relinquish their roles as programmable input/output pin to become in-system programming pins. A state machine controls the programming steps. The in-system programming structure can be cascaded in a 'daisy chain' fashion.

Growing community of inventors

Fremont, CA, United States of America

Chan-Chi Jason Cheng