초록 ▼

An integrated circuit which has two separate paths for two different data widths. The first processing path processes data up to n bits in a n multiplier. A second path operates in parallel with the first path, and includes smaller units which process data up to n 2 bits. The two paths can operate i

An integrated circuit which has two separate paths for two different data widths. The first processing path processes data up to n bits in a n multiplier. A second path operates in parallel with the first path, and includes smaller units which process data up to n 2 bits. The two paths can operate in parallel, but since the two paths have different data widths, they can more effectively operate with the different data sizes.

대표청구항 ▼

1. An integrated circuit comprising: a first processing path which processes data up to n-bits in an n-bit multiplier;a second processing path operating in parallel with said first path, including a plurality of arithmetic logic units which processes different data than the data in said first proces

1. An integrated circuit comprising: a first processing path which processes data up to n-bits in an n-bit multiplier;a second processing path operating in parallel with said first path, including a plurality of arithmetic logic units which processes different data than the data in said first processing path up to (n/2)-bits in parallel with said first processing path; anda first multiplexer which controls providing of n bit data to said n bit multiplier, and a second multiplexer separate from said first processing path which controls providing n/2 bit data to said second processing path.2. The digital signal processor of claim 1, wherein the first processing path processes data up to 16 bits.3. The digital signal processor of claim 2, wherein the second processing path processes data up to 8 bits.4. The digital signal processor of claim 1, wherein data of (n/2)-bits or less is processed in the second processing path.5. A method of processing data comprising: determining a size of the data to be processed; andselecting a first processing path with an n-bit multiplier if the data size is greater than a first size;selecting a second processing path with arithmetic logic units if the data size is less than the first size; andoperating said first and second processing paths in parallel.6. The method of claim 5, further comprising storing a result from the first processing path in a flop.7. The method of claim 5, further comprising storing a result from the second processing path in a flop.8. The method of claim 6, further comprising adding the result from the first processing path to a running total in at least one arithmetic logic unit.9. The method of claim 7, further comprising adding the result from the second processing path to a running total in at least one arithmetic logic unit.10. The method of claim 5, wherein the first processing path is used for approximately n-bit data and the second processing path is used for approximately (n/2)-bit data.