Electronic digital logic circuitry – Multifunctional or programmable – Array
Reexamination Certificate
1996-03-29
2001-09-18
Tokar, Michael (Department: 2819)
Electronic digital logic circuitry
Multifunctional or programmable
Array
C326S039000, C326S040000, C326S047000
Reexamination Certificate
active
06292018
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a configurable cellular array of dynamically configurable logic elements, such arrays being generally known as Field Programmable Gate Arrays (FPGAs).
BACKGROUND OF THE INVENTION
Reprogrammable FPGAs have been available commercially for several years. The best known commercial family of FPGAS are those from Xilinx Inc. One class of these devices uses Static Random Access Memory (SRAM) to hold control bits which control their configurations. Most FPGA devices replace traditional mask programmed Applications Specific Integrated Circuit (ASIC) parts which have a fixed configuration. The configuration of the FPGA is static and is loaded from a non-volatile memory when power is applied to the system. Nearly all commercially available FPGAs have a stream-based interface to the control store. (The control store contains the set of bits which determine what functions the FPGA will implement.) In a stream-based interface to the control store, a sequence of data is applied to a port in the FPGA to provide a complete configuration for the whole device or for a fixed (normally large) sub-section of the FPGA. This stream-based interface, when combined with an address counter which is implemented on the FPGA itself, provides an efficient method of loading the complete device configuration from adjacent EPROM or other non-volatile memory and power up without any additional overhead circuits. A stream based interface with an address counter is a suitable programming interface for an FPGA which is used as a replacement for a standard ASIC. Some FPGAs can be partly or totally reconfigured using one of a set of static configurations stored at different addresses in an EPROM, and can trigger the reconfiguration from within the design implemented on the FPGA.
Published International Application WO 90/11648, corresponding to U.S. Pat. No. 5,243,238, discloses an architecture hereafter referred to as CAL I, which has been implemented in an Algotronix product designated CAL 1024. CAL I is different from other commercially available FPGAs in that its control store appears as a standard SRAM to the systems designer, and can be accessed using address bus, data bus, chip enable, chip select and read/write signals. Addressing the control store as an SRAM supports a user program running on the host processor mapping the FPGA control store (configuration memory) into the memory or address space of the host processor so that the processor can configure the FPGA to implement a user-defined circuit. This arrangement, which is implemented in the CAL 1024 FPGA, allows the user to partition an application between the processor and the FPGA with appropriate sections being implemented by each device. The control store interface provides an important input/output (I/O) channel between the FPGA and the processor, although the I/O can also take place using more traditional techniques via, for example, a shared data memory area. This latter type of FPGA provides a passive control store interface because an external agent is required to initiate configuration or reconfiguration of the device, as required.
Experience with the CAL I architecture and trends within the electronics industry have made this second passive form of control store interface increasingly attractive for many applications. Microprocessors or microcontrollers are now pervasive components of computer systems and most board level systems contain one. The major benefit of the stream based “active” FPGA programming approach is that no overhead circuits are required to initiate reconfiguration. In systems where a microprocessor or microcontroller is present, the “passive” RAM emulating FPGA interface is preferable for several reasons:
(1) the FPGA configuration can be stored in the microprocessor's program and data memory (reducing the number of parts by removing the need for a separate memory chip),
(2) the existing data and address buses on the board can be used to control the FPGA (saving printed circuit board area by removing dedicated wires between the configuration EPROM and the FPGA);
(3) the FPGA control store can be written to and read from by the microprocessor, and thereby used as an I/O channel between the FPGA and the microprocessor, thereby potentially saving additional wiring between the FPGA and the processor buses and freeing the FPGA programmable I/O pins for communication with external devices, and
(4) the intelligence of the microprocessor can be used to support compression schemes for the configuration data and other techniques, which allows more flexibility in reprogramming the FPGA.
In addition, the difference in cost between an “active” FPGA with an associated EPROM holding its configuration and a passive FPGA with an active microcontroller chip containing an EPROM and a simple processor is minimal. The easy reprogrammability makes the passive FPGA attractive, even if the microcontroller has no other function apart from reprogramming the FPGA.
Another trend within the Electronics Industry has been the provision of “support chips” for microprocessors which provide an interface between I/O devices and a particular microprocessor. Examples of these devices include Universal Asynchronous Receiver Transmitters (UARTs) for low bandwidth serial I/O, Programmable Peripheral Interfaces (PPIs) for low bandwidth parallel I/O and various specialised chips for higher bandwidth connections to networks and disk drives. These support chips appear to the processor as locations in the I/O or memory address space to and from which data are transferred. Some support chips can interrupt the processor via interrupt lines or take over the bus for Direct Memory Access (DMA) operations. In many ways a passive FPGA chip can be viewed as a successor to a support chip, providing an interface to the processor via its control store on the one hand an interface to the external world via a large number of flexible I/O lines on the other, for example 128 programmable I/O lines on the Algotronix CAL 1024 device.
A passive FPGA chip has a number of advantages. For example, it is cost-effective to provide a single FPGA with a library of configurations instead of providing a number of support chips. In addition, providing a single FPGA for several functions reduces the number of devices in the processor manufacturer's catalogue. Also, reconfigurable FPGAs can support changeable I/O functions, such as when a single external connector can be used as either a serial or a parallel port. With a passive RAM control interface, the FPGA is able to support other functions as well.
Each time an FPGA is reconfigured to implement a different set of functions, the microprocessor must access the configuration memory. One reconfiguration typically requires many control store accesses, one access for each word of configuration memory to be changed. Several important classes of reconfiguration have been identified.
(1) Application swapping occurs when one application terminates and a completely different application wishes to make use of the FPGA. In this case the FPGA chip is completely reconfigured, usually from a static configuration.
(2) Task swapping occurs when the application must configure relatively large sections of the FPGA to implement a new phase in the computation. For example, a sorting application might first sort small batches of data completely using configuration A and then merge those sorts into a completely sorted stream of data using configuration B. In this case, the application has knowledge of both configurations and need only change those resources which are different in configuration B. At a later point, configuration A may itself be restored.
(3) Data dependent reconfiguration occurs when the configurations of some cells are computed dynamically based on input data by the application program, rather than being loaded from a static configuration file. Often a static configuration is first loaded, then a relatively small sub-set of cells are reconfigured dynamically (that is, reconfigured while
Casey Michael R.
Oblon & Spivak, McClelland, Maier & Neustadt P.C.
Paik Steven S.
Tokar Michael
Xilinx , Inc.
LandOfFree
Configurable cellular array does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Configurable cellular array, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Configurable cellular array will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2465269