Method for configuring a finite impulse response filter in a programmable logic device

Abstract

A hybrid FIR filter includes a plurality of FIR filter units arranged as Direct Form FIR filters, connected together in an arrangement similar to a Transpose Form FIR filter. The hybrid filter arrangement may be used to configure a larger FIR filter in a programmable logic device having one or more specialized functional blocks, incorporating multipliers and adders, that are particularly well-suited for configuration as small Direct Form FIR filters.

Description

BACKGROUND OF THE INVENTION

This invention relates to a method for configuring a finite impulse response filter in a programmable logic device, and more particularly to efficiently configuring a finite impulse response filter of arbitrary size.

Programmable logic devices are well known. Early programmable logic devices were one-time configurable. For example, configuration may have been achieved by “blowing”—i.e., opening—fusible links. Alternatively, the configuration may have been stored in a programmable read-only memory. Those devices generally provided the user with the ability to configure the devices for “sum-of-products” (or “P-TERM”) logic operations. Later, such programmable logic devices incorporating erasable programmable read-only memory (EPROM) for configuration became available, allowing the devices to be reconfigured.

Still later, programmable logic devices incorporating static random access memory (SRAM) elements for configuration became available. These devices, which also can be reconfigured, store their configuration in a nonvolatile memory such as an EPROM, from which the configuration is loaded into the SRAM elements when the device is powered up. These devices generally provide the user with the ability to configure the devices for look-up-table-type logic operations.

As programmable logic devices have become larger, it has become more common to add specialized blocks to perform particular functions that have become more common in programmable logic devices. For example, at some point, such devices began to be provided with embedded blocks of random access memory that could be configured by the user to act as random access memory, read-only memory, or logic (such as P-TERM logic); such embedded blocks might even be provided in different sizes on the same device. Other types of memory, such as read-only memory (ROM) or shift registers, also have been provided. More recently, multiplier circuits have been provided on programmable logic devices. Whereas in prior programmable logic devices space was not available for dedicated multipliers, current larger devices can accommodate multipliers. This spares users from having to create multipliers by configuring the available logic. Moreover, as described in commonly-assigned U.S. Pat. No. 6,538,470, which is hereby incorporated by reference in its entirety, specialized multiplier blocks may be provided including multipliers and other arithmetic circuits such as adders and/or subtracters and/or accumulators. Such blocks are sometimes referred to as “multiplier-accumulator blocks” or “MAC blocks.” Such blocks, for example, may be useful in digital signal processing, such as is performed in audio applications, and therefore such specialized multiplier blocks also are sometimes referred to as “DSP blocks.”

Such specialized multiplier blocks typically are capable of operations up to a certain size. For example, a specialized multiplier block may be provided that can perform a single 36-bit-by-36-bit multiplication operation, or up to four individual 18-bit-by-18-bit multiplication operations which can be combined by the aforementioned adders.

One use for such a specialized multiplier block may be as a finite impulse response (FIR) filter, or portion of a FIR filter. In a FIR filter, which is commonly used for DSP operations, samples of a signal to be processes are multiplied by a fixed set of coefficients and those products are added together. A FIR filter may be characterized by its number of taps, which corresponds to the number of multipliers and also to the number of coefficients.

A specialized multiplier block of the type described above necessarily has a finite number of multipliers. While such a block is particularly well-adapted to be configured as a FIR filter, if the number of multipliers required (i.e., the number of taps) exceeds the number of multipliers in the block, then more than one block must be used to create the required filter. For example, specialized multiplier blocks in the STRATIX® family of PLDs available from Altera Corporation, of San Jose, Calif., specialized multiplier blocks typically have four multipliers. On the other hand, in many DSP applications, FIR filters with over 200 taps are not unheard of. Thus, implementation of a FIR filter is frequently spread over a large number of specialized multiplier blocks, giving rise to routing and speed issues.

It would be desirable to be able to configure a FIR filter in a programmable logic device as efficiently as possible for a given number of filter taps.

SUMMARY OF THE INVENTION

The present invention achieves greater efficiency in the configuration of specialized functional blocks of the type described above as FIR filters, by providing a configuration method that combines aspects of different forms of FIR filters in a way that is more efficient.

FIR filters can be arranged in any of several forms. Among those forms are the Direct Form (including Direct Form I and Direct Form II) and the Transpose Form. In the Direct Form, a set of coefficients is input to the multipliers, and the entire chain of samples is shifted to each multiplier. On each clock cycle, all of the products are different, and are summed to provide the output for that clock cycle. Although Direct Form FIR filters map easily onto specialized functional blocks of the type described above—indeed, a 4-tap Direct Form FIR filter can be implemented entirely in one such block in the aforementioned STRATIX® family of devices—larger filters may require a long input sample chain and a large adder tree that is difficult to route and limits the maximum clock speed attainable.

In a Transpose Form FIR filter, the coefficients are input as above. On each clock cycle, the same sample is input to all multipliers. The result of each multiplication is stored in a one-cycle delay and added on the next cycle to the result of the adjacent multiplication. The output for each cycle is the sum at the end of the chain. A Transpose Form FIR filter therefore requires no sample chain and no adder tree. However, the required adders become progressively wider down the chain (because the sum becomes wider as more stages are added in), with the result that a large number of adders have large widths, which limits the maximum clock speed attainable. In addition, because each stage requires one multiplier and one adder (to add each product to the delayed sum of the upstream stages), Transpose Form FIR filters do not map well onto specialized functional blocks of the type described above, which generally have fewer adders than multipliers.

In accordance with the invention, specialized functional blocks in PLDs are configured as a hybrid form of FIR filter, including characteristics of both Direct Form and Transpose Form FIR filters, but providing mathematically identical results and having fewer disadvantages than either form when mapped onto specialized functional blocks. The invention provides a hybrid FIR filter including a first number of Transpose Form FIR filter stages, where each stage is a Direct Form FIR filter.

A method is provided according to the invention for programmably configuring a programmable logic device as a hybrid FIR filter, where the programmable logic device is of a type having a plurality of specialized processing blocks each of which includes a plurality of multipliers and circuitry for adding outputs of the multipliers. The method includes programmably configuring each of a first number of the specialized processing blocks as a respective Direct Form FIR filter. An adder chain is programmably configured to add outputs of the Direct Form FIR filters. The adder chain includes a respective delay between each of the Direct Form FIR filters. As a result, the programmable logic device is configured as a Transpose Form FIR filter having a number of stages equal to said first number, with each stage being one of the Direct Form FIR filters.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other advantages of the invention will be apparent upon consideration of the following detailed description, taken in conjunction with the accompanying drawings, in which like reference characters refer to like parts throughout, and in which:

FIG. 1 is a schematic representation of a 4-tap Direct Form FIR filter;

FIG. 2 is a schematic representation of a 4-tap Transpose Form FIR filter;

FIG. 3 is a representation of the chaining of results between stages of a Transpose Form FIR filter;

FIG. 4 is a schematic representation of a multitap hybrid form FIR filter in accordance with the present invention, incorporating stages each of which is a 4-tap Direct Form FIR filter;

FIG. 5 is a schematic representation of a multitap hybrid form FIR filter in accordance with the present invention, incorporating stages each of which is a Direct Form FIR filter of generic width;

FIG. 6 is a cross-sectional view of a magnetic data storage medium encoded with a set of machine-executable instructions for performing the method according to the present invention; and

FIG. 7 is a cross-sectional view of an optically readable data storage medium encoded with a set of machine executable instructions for performing the method according to the present invention.

DETAILED DESCRIPTION OF THE INVENTION

As described above, FIR filters are common structures used in many DSP applications. Mathematically, a FIR filter may be described as:

$\mathrm{Yk}=\sum _{i=0}^{\mathrm{Taps}-1}\mathrm{Ci}•\mathrm{Sk}-i$ where Y_k is the kth output term, c_i is the ith coefficient, s_k-i is the (k−i)th sample, and Taps is the number of taps in the filter.

The structure of a 4-tap Direct Form II FIR filter 10 is shown in FIG. 1, and includes four multipliers 11. The multiplier outputs are added by adder tree 12, shown schematically in collapsed form, to provide term Y_k. Each multiplier 11 has a coefficient input 13 for one of the four coefficients c₀-c₃. In some applications, the coefficients are held constant, while in others, such as adaptive filters, decimation filters, interpolation filters, etc., they may change. Each multiplier 11 also has a sample input 14, and the samples s_k are clocked in via sample chain 15. The kth term Y_k is provided when coefficient c₀ is being multiplied by the kth sample s_k and the other coefficients c₁-c₃ are being multiplied by the (k−1)th through (k−3)th samples, added by adder 16 as shown. Thus, as each new sample is clocked in, the previous samples are reused, so that they are used a total number of times equal to the number of taps as follows:y₀=c₀s₀+c₁s₋₁+c₂s₋₂+c₃s₋₃ y₁=c₀s₁+c₁s₀+c₂s₋₁+c₃s₋₂ y₂=c₀s₂+c₁s₁+c₂s₋₀+c₃s₋₁ y₃=c₀s₃+c₁s₂+c₂s₁+c₃s₀ y₄=c₀s₄+c₁s₃+c₂s₂+c₃s₁

As explained above, if the number of taps is large, the number of multipliers needed is large, as is the size of the required adder tree, and moreover the sample rate dictates the number of multiplications that must be performed during each clock cycle. For more than four taps, sample chain 15 would extend to the right, to the next group of multipliers. The number of taps that may be required in any application is arbitrary. Therefore, in a PLD implementation, where the number of multipliers and associated adders grouped in a single block is necessarily finite, a large number of taps translates into a large number of blocks, leading to a large routing problem as well as the implementation of a large adder tree, with all of the consequent effects on latency and attainable clock speeds. At the same time, however, the structure of the specialized functional blocks described above is well-suited to the Direct Form II FIR filter.

FIG. 2 shows the structure of a 4-tap Transpose Form FIR filter 20, which again includes four multipliers 11. Once again, each multiplier 11 has a coefficient input 13, and the four coefficients c₀-c₃ are held constant. However, instead of clocking in the samples s_k via a sample chain as in FIG. 1, in Transpose FIR filter 20, the kth sample is input to each multiplier 11 substantially simultaneously. In other words, all multipliers 11 receive the same sample input, with the output of each multiplier added by respective adder 22 to the delayed output of the previous adder 22, as delayed by one of one-clock-cycle delay elements 23. Each combination of a multiplier 11, adder 22 and delay element 23 may be referred to as a “stage” 24 of filter 20. In the first stage, the second input to adder 22 may be zero, as shown. The kth term Y_k is provided from the stage 24 whose coefficient is c₀. Thus, only one sample is required at any one time, but the results for each sample are stored in delays 23 and added into the next stage, as shown in FIG. 3.

The Transpose Form FIR filter has the advantage, as compared to the Direct Form, that neither a sample chain nor a large adder tree is required. However, the average adder width is greater than in the Direct Form. Indeed, in some embodiments, as many as half of all adders used to implement the filter are of the maximum width. Large adder widths have a detrimental effect on the maximum attainable clock speed. In addition, as compared to the Direct Form FIR filter, a Transpose Form FIR filter is not as easily implemented in the specialized functional block described above.

The present invention combines the advantages of the Direct Form and Transpose Form FIR filters. In accordance with the invention, one embodiment of which is shown in FIG. 4, a FIR filter 40 of the Transpose Form type has as each of its stages 41 a FIR filter unit 42 of the Direct Form type. In this embodiment 40, FIR filter unit 42 includes four multipliers 420. The specialized multiplier blocks described above, having four multipliers, are particularly well-suited to implementing this embodiment 40.

Instead of inputting one sample s_k at a time to all stages, four samples s_t are input at any one time. The most recent sample s_t is input substantially simultaneously to the leftmost (as depicted in FIG. 4) multiplier 420 in each of units 42 while each of three less recent samples s_t is input substantially simultaneously to a respectively further rightward (as depicted in FIG. 4) one of the multipliers 420. A chain 43 of registers or delays 430 of one clock period each is required to maintain the four samples s_t.

The number—in this embodiment, four—of delays 430 in chain 43 may be referred to as the depth, n_d, of the hybrid Transpose/Direct Form FIR filter 40. A cascade adder chain 44 is used to add the results of the stages 41. Unlike Transpose Form FIR filter 20 in which the output of each stage is stored in a one-clock-cycle delay 23, because it takes four clock periods to clock all samples into sample chain 43, each delay 441 between adders 440 of adder chain 44 is four clock periods long. The number of stages 41 of hybrid filter 40 may be referred to as the width, n_t, of filter 40. The product n_t×n_d is equal to the number of taps of filter 40.

Similarly to the case of Transpose Form FIR filter 20, in which only one sample is required at any one time, but the results for each sample are stored in delays 23 and added into the next stage, as shown in FIG. 3, in hybrid Transpose/Direct Form FIR filter 40, only four samples are required at any one time, but the results for each group of four samples are stored in delays 441 and added into the next stage.

FIG. 5 shows the hybrid Transpose/Direct Form FIR filter according to the invention, generalized to arbitrary n_d and arbitrary n_t. In embodiment 50, each stage 51 is a FIR filter unit 52 of the Direct Form type, including an arbitrary number n_d of multipliers 520. Similarly, sample chain 53 includes n_d one-clock-cycle delays 530, and adder chain 54, used to add the results of the stages 51, includes delays 541 between adders 540, with each delay 541 being n_d clock periods long.

It will be apparent that for n_d=1, the Transpose/Direct Form FIR filter according to the invention collapses to an ordinary Transpose Form FIR filter, while for n_t=1, the Transpose/Direct Form FIR filter according to the invention collapses to an ordinary Direct Form FIR filter.

Although the sample chain required for the Transpose/Direct Form FIR filter according to the invention is longer than that needed for any ordinary Transpose Form FIR filter, it is nevertheless shorter than that needed for a Direct Form FIR filter of arbitrary length, and is limited to a finite length of n_d. The hybrid structure maps well onto specialized processing blocks having a finite number of multipliers. In particular, for the case 40 of n_d=4, shown in FIG. 4, the hybrid structure maps well onto the aforementioned specialized processing blocks. But in any case, regardless of the particular type of specialized processing block, there will be a value of n_d for which the hybrid structure maps well. Moreover, because of the delays inherent in adder chain 44/54, the adder chain can be split over several clock cycles, with minimal impact on resources even for large adders.

It is common in specialized multiplier blocks to include pipelining registers, e.g., before and after the multipliers, and between different adder stages. Those pipelining registers, if provided, can be counted towards the total number of delays (z^−nd). This would lead to a more efficient implementation of the invention.

The hybrid Transpose/Direct Form FIR filter according to the invention can be used in different kinds of integrated circuit devices including, without limitation, application-specific integrated circuits and PLDs, and the invention includes the method of programming the hybrid Transpose/Direct Form FIR filter into the specialized multiplier block of a PLD. PLDs programmed in accordance with the invention may be used in many kinds of electronic devices. One possible use is in a data processing system that also may include one or more of the following components: a processor; memory; I/O circuitry; and peripheral devices. These components may be coupled together by a system bus and are populated on a circuit board which is contained in an end-user system.

The end user system can be used in a wide variety of applications, such as computer networking, data networking, instrumentation, video processing, digital signal processing, or any other application where the advantage of using programmable or reprogrammable logic is desirable. The PLD can be used to perform a variety of different logic functions. For example, the PLD can be configured as a processor or controller that works in cooperation with the main system processor. The PLD may also be used as an arbiter for arbitrating access to shared resources in the system. In yet another example, the PLD can be configured as an interface between the processor and one of the other components in the system.

Various technologies can be used to implement integrated circuit devices of the type described above, including PLDs that can be programmed according to this invention.

Instructions for carrying out the method according to this invention may be encoded on a machine-readable medium, to be executed by a suitable computer or similar device to implement the method of the invention for programming PLDs. For example, a personal computer may be equipped with an interface to which a PLD can be connected, and the personal computer can be used by a user to program the PLD using a suitable software tool, such as the QUARTUS® II software described above.

FIG. 6 presents a cross section of a magnetic data storage medium 600 which can be encoded with a machine executable program that can be carried out by systems such as the aforementioned personal computer, or other computer or similar device. Medium 600 can be a floppy diskette or hard disk, or magnetic tape, having a suitable substrate 601, which may be conventional, and a suitable coating 602, which may be conventional, on one or both sides, containing magnetic domains (not visible) whose polarity or orientation can be altered magnetically. Except in the case where it is magnetic tape, medium 600 may also have an opening (not shown) for receiving the spindle of a disk drive or other data storage device.

The magnetic domains of coating 602 of medium 600 are polarized or oriented so as to encode, in manner which may be conventional, a machine-executable program, for execution by a programming system such as a personal computer or other computer or similar system, having a socket or peripheral attachment into which the PLD to be programmed may be inserted, to configure appropriate portions of the PLD, including its specialized processing blocks, if any, as a hybrid Transpose Form/Direct Form FIR filter in accordance with the invention.

FIG. 7 shows a cross section of an optically-readable data storage medium 700 which also can be encoded with such a machine-executable program, which can be carried out by systems such as the aforementioned personal computer, or other computer or similar device. Medium 700 can be a conventional compact disk read only memory (CD-ROM) or digital video disk read only memory (DVD-ROM) or a rewriteable medium such as a CD-R, CD-RW, DVD-R, DVD-RW, DVD+R, DVD+RW, or DVD-RAM or a magneto-optical disk which is optically readable and magneto-optically rewriteable. Medium 700 preferably has a suitable substrate 701, which may be conventional, and a suitable coating 702, which may be conventional, usually on one or both sides of substrate 701.

In the case of a CD-based or DVD-based medium, as is well known, coating 702 is reflective and is impressed with a plurality of pits 703, arranged on one or more layers, to encode the machine-executable program. The arrangement of pits is read by reflecting laser light off the surface of coating 702. A protective coating 704, which preferably is substantially transparent, is provided on top of coating 702.

In the case of magneto-optical disk, as is well known, coating 702 has no pits 703, but has a plurality of magnetic domains whose polarity or orientation can be changed magnetically when heated above a certain temperature, as by a laser (not shown). The orientation of the domains can be read by measuring the polarization of laser light reflected from coating 702. The arrangement of the domains encodes the program as described above.

It will be understood that the foregoing is only illustrative of the principles of the invention, and that various modifications can be made by those skilled in the art without departing from the scope and spirit of the invention, and the present invention is limited only by the claims that follow.

Claims

1. A programmable logic device configured as hybrid finite impulse response (FIR) filter circuitry; wherein: said programmable logic device has a plurality of specialized processing blocks each of which includes a plurality of multipliers and a single adder for simultaneously adding outputs of said multipliers; andsaid hybrid FIR filter circuitry comprises:a first number of said specialized processing blocks each of which is configured as a respective Direct Form FIR filter circuit in which each respective multiplier in said plurality of multipliers multiplies a respective sample by a respective coefficient and outputs of said respective multipliers are added simultaneously by said single adder to generate a Direct Form FIR output, said hybrid FIR filter circuitry having a second number of taps, each said Direct Form FIR filter circuit comprising a third number of taps, and said second number being a product of said first number and said third number;an adder chain comprising a plurality of additional adders, separate from any adder in said first number of said specialized processing blocks that adds said outputs of said respective multipliers, each of said additional adders having, as one input, one output of one of said respective Direct Form FIR filter circuits, and each of said additional adders other than a first of said additional adders having as one input an output of another one of said additional adders, said adder chain further comprising a respective output delay between (a) each said output of another one of said additional adders, that is an input to one of said additional adders other than a first one of said additional adders, and (b) said one of said additional adders other than a first one of said additional adders, andan input sample chain comprising a number of registers equal to said third number; wherein:each said register provides a sample substantially simultaneously to a corresponding input of each said Direct Form FIR filter circuit,said input sample chain shifts once at every clock cycle,each Direct Form FIR filter circuit generates an output at every said clock cycle, andeach said respective output delay has a delay period of a number of said clock cycles equal to said third number.

2. The programmable logic device of claim 1 wherein: each said Direct Form FIR filter circuit comprises a number of multiplier circuits equal to said third number; andeach said corresponding input comprises an input to a respective one of said multiplier circuits in each said Direct Form FIR filter circuit.

3. The programmable logic device of claim 2; wherein: said hybrid FIR filter circuitry further comprises a number of coefficients equal to said second number;said coefficients are divided into a number of groups equal to said first number;each said group comprises a number of coefficients equal to said third number; andeach coefficient in each said group is input to a respective one of said multiplier circuits.

4. The programmable logic device of claim 1 wherein: each said Direct Form FIR filter circuit comprises a number of multiplier circuits equal to said third number;said hybrid FIR filter circuitry further comprises a number of coefficients equal to said second number;said coefficients are divided into a number of groups equal to said first number;each said group comprises a number of coefficients equal to said third number; andeach coefficient in each said group is input to a respective one of said multiplier circuits.

5. A method of programmably configuring a programmable logic device as a hybrid finite impulse response (FIR) filter, said programmable logic device having a plurality of specialized processing blocks each of which includes a plurality of multipliers and a single adder for simultaneously adding outputs of said multipliers, said method comprising: programmably configuring each of a first number of said specialized processing blocks as a respective Direct Form FIR filter in which each respective multiplier in said plurality of multipliers multiplies a respective sample by a respective coefficient and outputs of said respective multipliers are added simultaneously by said single adder to generate a Direct Form FIR output, said hybrid FIR filter having a second number of taps, each said respective Direct Form FIR filter comprising a third number of taps, and said second number being a product of said first number and said third number;programmably configuring an adder chain from a plurality of additional adders, separate from any adder in said first number of said specialized processing blocks that adds said outputs of said respective multipliers, to add outputs of said Direct Form FIR filters, said adder chain comprising a respective delay preceding each of said additional adders in said adder chain, other than a first one of said additional adders in said adder chain, that has, as an input, one of said outputs of said Direct Form FIR filters;programmably configuring each said respective delay to have a delay period of a number of clock cycles equal to said third number; andprogrammably configuring an input sample chain comprising a number of registers equal to said third number; wherein:each said register provides a sample substantially simultaneously to a corresponding input of each said Direct Form FIR filter,said input sample chain shifts once at every clock cycle,each Direct Form FIR filter circuit generates an output at every said clock cycle, andsaid programmable logic device is configured as a hybrid FIR filter having a number of hybrid stages equal to said first number, each of said hybrid stages comprising a respective one of said Direct Form FIR filters.

6. The method of claim 5 wherein: each said Direct Form FIR filter comprises a number of multipliers equal to said third number; andeach said corresponding input comprises an input to a respective one of said multipliers in each said Direct Form FIR filter.

7. The method of claim 6 wherein said hybrid FIR filter comprises a number of coefficients equal to said second number; said method further comprising: programmably configuring a number of coefficient memories equal to said first number, each said coefficient memory storing a number of coefficients equal to said third number; andprogrammably configuring each coefficient stored in each said memory as an input to a respective one of said multipliers.

8. The method of claim 5 wherein: said programmably configuring each of a first number of said specialized processing blocks as a respective Direct Form FIR filter comprises programmably configuring each said Direct Form FIR filter to have a number of multipliers equal to said third number; andsaid hybrid FIR filter further comprises a number of coefficients equal to said second number; said method further comprising:programmably configuring a number of coefficient memories equal to said first number, each said coefficient memory storing a number of coefficients equal to said third number; andprogrammably configuring each coefficient stored in each said memory as an input to a respective one of said multipliers.

9. A non-transitory data storage medium encoded with non-transitory machine-executable instructions for performing a method of programmably configuring a programmable logic device as a hybrid finite impulse response (FIR) filter, said programmable logic device having a plurality of specialized processing blocks each of which includes a plurality of multipliers and a single adder for simultaneously adding outputs of said multipliers, said instructions comprising: instructions to programmably configure each of a first number of said specialized processing blocks as a respective Direct Form FIR filter in which each respective multiplier in said plurality of multipliers multiplies a respective sample by a respective coefficient and outputs of said respective multipliers are added simultaneously by said single adder to generate a Direct Form FIR output, said hybrid FIR filter having a second number of taps, said instructions to programmably configure each of said first number of said specialized processing blocks as said respective Direct Form FIR filter comprises instructions to programmably configure each said Direct Form FIR filter with a third number of taps, said second number being a product of said first number and said third number;instructions to programmably configure an adder chain from a plurality of additional adders, separate from any adder in said first number of said specialized processing blocks that adds said outputs of said respective multipliers, to add outputs of said Direct Form FIR filters, said adder chain comprising a respective delay preceding each of said additional adders in said adder chain, other than a first one of said additional adders in said adder chain, that has as an input one of said Direct Form FIR filters;instructions to programmably configure each said respective delay to have a delay period of a number of clock cycles equal to said third number; andinstructions to programmably configure an input sample chain comprising a number of registers equal to said third number; wherein:each said register provides a sample substantially simultaneously to a corresponding input of each said Direct Form FIR filter,said input sample chain shifts once at every clock cycle,each Direct Form FIR filter circuit generates an output at every said clock cycle, andsaid programmable logic device is configured as a hybrid FIR filter having a number of hybrid stages equal to said first number, each of said hybrid stages comprising a respective one of said Direct Form FIR filters.

10. The non-transitory data storage medium of claim 9 wherein said instructions further comprise: instructions to configure each said Direct Form FIR filter with a number of multipliers equal to said third number; wherein:each said corresponding input comprises an input to a respective one of said multipliers in each said Direct Form FIR filter.

11. The non-transitory data storage medium of claim 10 wherein said hybrid FIR filter comprises a number of coefficients equal to said second number; said instructions further comprising: instructions to programmably configure a number of coefficient memories equal to said first number, each said coefficient memory storing a number of coefficients equal to said third number; andinstructions to programmably configure each coefficient stored in each said memory as an input to a respective one of said multipliers.

12. The non-transitory data storage medium of claim 9 wherein: said instructions to programmably configure each of a first number of said specialized processing blocks as a respective Direct Form FIR filter comprise instructions to programmably configure each said Direct Form FIR filter to have a number of multipliers equal to said third number; andsaid hybrid FIR filter further comprises a number of coefficients equal to said second number; said instructions further comprising:instructions to programmably configure a number of coefficient memories equal to said first number, each said coefficient memory storing a number of coefficients equal to said third number; andinstructions to programmably configure each coefficient stored in each said memory as an input to a respective one of said multipliers.