Abstract:
A device reduces false positive memory error detections by using a masking unit and sensitivity mask data to exclude unused portions of the memory from the error detection computations. A device includes an error detection unit to read data from the memory and verify data integrity. The sensitivity mask data indicates unused portions of the memory. Unused portions of the memory may correspond with configuration data for unused portions of a programmable device. Each bit of the sensitivity mask data may indicate the usage of one or more bits of the data from the memory. In response to the mask data, the masking unit sets data from the unused portions of the memory to values that do not change the result of the error detection computations. This prevents any errors in data from the unused portions of the memory from raising an error signal.
Abstract:
A programmable logic device has a plurality of super-regions of programmable logic disposed on the device in a two-dimensional array of intersecting rows and columns of super-regions. Horizontal and vertical inter-super-region interconnection conductors are associated with each row and column, respectively. Each super-region includes a plurality of regions of programmable logic, and each region includes a plurality of subregions of programmable logic. Inter-region interconnection conductors are associated with each super-region, principally for bringing signals into the super-region and interconnecting the regions in the super-region. Local conductors are associated with each region, principally for bringing signals into the region. At the super-region level the device may be horizontally and vertically isomorphic, which helps make it possible to produce devices with low aspect ratios of one or nearly one. Shared driver circuits may be provided (e.g., for (1) receiving signals from the subregions and the horizontal and/or vertical conductors, and (2) applying selected received signals to the inter-region conductors, the horizontal and vertical conductors, and possibly also the local conductors). The horizontal and/or vertical conductors may be axially segmented and buffering circuitry may be provided for programmably stitching together axial segments to make longer conductors.
Abstract:
A programmable logic array integrated circuit device includes a plurality of regions of programmable logic disposed on the device in a two-dimensional array of intersecting rows and columns. Interconnection conductors are associated with each row and column. The interconnection conductors associated with each row include some that extend continuously along the entire length of the row and some that extend continuously along only the left or right half of the row. To increase the flexibility with which the logic regions can be connected to the row and column conductors, adjacent regions are paired and circuitry is provided for allowing the outputs of each pair to be swapped for driving the row and column conductors. Registers in logic regions can still be used for other purposes when not being used to register the main combinatorial outputs of the logic regions. Many other enhanced features are also provided.
Abstract:
A programmable logic array integrated circuit device has a plurality of regions of programmable logic disposed on the device in a two-dimensional array of intersecting rows and columns of regions. The output signals of several regions share a group of drivers for applying region output signals to interconnection conductors that convey signals between regions. This conserves driver resources and increases signal routing flexibility. Various approaches can be used for configuring the interconnection conductors to also conserve interconnection conductor resources. Logic regions may be used to directly drive specific input/output cells, thereby simplifying signal routing to the I/O cells and also possibly simplifying the structure of the I/O cells (e.g., by allowing certain I/O cell functions to be performed in the associated logic region). Region output signal routing flexibility may also be enhanced to facilitate simultaneous performance of combinatorial logic and a separate “lonely register” function in modules of the regions.
Abstract:
A programmable logic device is provided that allows a redundant row of programmable logic to be shifted into place to repair the device when a defect is detected in a row of programmable logic on the device. The redundant row is shifted into place by routing programming data into the normal logic and the redundant logic while bypassing the row of logic containing the defect. Switching circuitry may be used to direct programming data into the serial inputs of various data registers that are then used to load the programming data into the device. The patterns of programmable connections that are made between programmable logic regions on the device and vertical and horizontal conductors also allow redundant logic to be shifted into place. Some connections between the logic and the horizontal and vertical conductors may be identical within a column to facilitate shifting. Other connections may only partially overlap between respective rows. The partially overlapping pattern allows the connections to be less regular, which increases flexibility when routing signals on the device.
Abstract:
In order to facilitate the performance of multiplications in programmable logic devices, individual logic modules of such devices are constructed so that one logic module can perform (at least) both one place of binary multiplication and one place of full binary addition. This makes it possible to reduce the number of logic modules that are required to perform a multiplication. It also reduces the number of inter-module connections employed in a multiplication, thereby tending to decrease the time required to perform a multiplication.
Abstract:
A programmable logic array integrated circuit device includes a plurality of regions of programmable logic disposed on the device in a two-dimensional array of intersecting rows and columns of such regions. Each row has a plurality of adjacent horizontal conductors, and each column has a plurality of adjacent vertical conductors. The regions in a row are interspersed with groups of local conductors which interconnect the adjacent regions and the associated horizontal and vertical conductors. The local conductors can also be used for intra-region communication, as well as communication between adjacent regions. Secondary signals such as clocks and clears for the regions can be drawn either from dedicated secondary signal conductors or normal region inputs. Memory cell requirements for region input signal selection are reduced by various techniques for sharing these memory cells.
Abstract:
A voltage translation circuit 6 and method of operating a circuit to provide voltage level translation, particularly for use in association with a programmable logic device 2. An input signal, is provided to an NMOS transistor N2 and a CMOS inverter 12. The inverter 12 drives an NMOS pulldown transistor N4. The drain of N2 is coupled to the input of an inverter 14, the output of which is also coupled to the translation circuit output. A source of PMOS pullup transistor P1 drives the input of inverter 14 to pullup the output to a high level.
Abstract:
The output buffer includes first and second N channel FETs serially connected between a ground terminal and a voltage supply terminal and having a common terminal connected to an output terminal. Third and fourth N channel FETs are serially connected between the ground terminal and the voltage supply terminal and have a common terminal connected to the output terminal. A first logic circuit responds to data input (DIN) signals and an operation enable bar (OEB) signal for applying a conductive bias voltage to the third N channel FET, and a second logic circuit responds to the DIN signals and the OEB signal for applying a conductive bias voltage to the fourth N channel FET. A first P channel FET couples the conductive bias voltage on the third N channel FET to the first N channel FET, and a second P channel FET couples the conductive bias on the fourth N channel FET to the second N channel FET. The P channel transistors are mask programmable weak conductors thus limiting the rate of conductive bias applied to the first N channel FET and to the second N channel FET. Fifth and sixth N channel FETs are connected between the gate terminals of the first and second FETs and circuit ground for preventing conduction of the first and second FETs when the fifth and sixth FETs are conductive.
Abstract:
A programmable logic device has a plurality of super-regions of programmable circuitry disposed on the device in a two-dimensional array of such super-regions. Each super-region includes a plurality of regions of programmable logic and a region of programmable memory. Each logic region includes a plurality of subregions of programmable logic. Each super-region has associated interconnection resources for allowing communication between the logic and memory regions of that super-region without the need to use, for such relatively local interconnections, the longer-length inter-super-region interconnection resources that are also provided on the device.