Abstract: A direct memory access (DMA) engine schedules data transfer requests of a data processing system according to both an assigned transfer priority and the deadline for completing a transfer.

Abstract: A size and retry programmable multi-synchronous FIFO. In one embodiment, a multi-synchronous FIFO memory generally comprises a selectable number of addressable memory locations for storing information; read control means synchronized by a read clock for controlling pop transactions configured to read from one or more of the selected number of addressable memory locations; write control means synchronized by a write clock asynchronous to the read clock for controlling push transactions to write to one or more of the selected number of addressable memory locations; and selectable transaction retry control means configured to cause read control means to repeat selected pop transactions and/or cause write control means to repeat selected push transactions.

Abstract: A direct memory access (DMA) engine schedules data transfer requests of a data processing system according to both an assigned transfer priority and the deadline for completing a transfer.

Abstract: A direct memory access (DMA) engine schedules data transfer requests of a data processing system according to both an assigned transfer priority and the deadline for completing a transfer.

Abstract: A global timebase system and method for a system-on-chip synchronizes multiple clock domains in each of a plurality of receiver modules by broadcasting a global timebase count value as Gray code over a global timebase bus. A global timebase generator includes a binary counter and a binary-to-Gray-code converter. Each receiver module registers the global timebase count value with its own local clock and includes a Gray-code-to-binary converter. The converted value, in binary form, may be used as least significant bits of a globally synchronized local timebase. Most significant bits may be generated by a local binary counter incremented at each 1-to-0 transition of the most significant bit of the global timebase count value.

Abstract: A direct memory access (DMA) engine schedules data transfer requests of a system-on-chip data processing system according to both an assigned transfer priority and the deadline for completing a transfer. Transfer priority is based on a hardness representing the penalty for missing a deadline. Priorities are also assigned to zero-deadline transfer requests in which there is a penalty no matter how early the transfer completes. If desired, transfer requests may be scheduled in timeslices according to priority in order to bound the latency of lower priority requests, with the highest priority hard real-time transfers wherein the penalty for missing a deadline is severe are given the largest timeslice. Service requests for preparing a next data transfer are posted while a current transaction is in progress for maximum efficiency. Current transfers may be preempted whenever a higher urgency request is received.

Abstract: A multiply accumulate unit (“MAC”) that performs operations on packed integer data. In one embodiment, the MAC receives 2 32-bit data words which, depending on the specified mode of operation, each contain either four 8-bit operands, two 16-bit operands, or one 32-bit operand. Depending on the mode of operation, the MAC performs either sixteen 8×8 operations, four 16×16 operations, or one 32×32 operation. Results may be individually retrieved from registers and the corresponding accumulator cleared after the read cycle. In addition, the accumulators may be globally initialized. Two results from the 8×8 operations may be packed into a single 32-bit register. The MAC may also shift and saturate the products as required.

Abstract: A size and retry programmable multi-synchronous FIFO. In one embodiment, a multi-synchronous FIFO memory generally comprises a selectable number of addressable memory locations for storing information; read control means synchronized by a read clock for controlling pop transactions configured to read from one or more of the selected number of addressable memory locations; write control means synchronized by a write clock asynchronous to the read clock for controlling push transactions to write to one or more of the selected number of addressable memory locations; and selectable transaction retry control means configured to cause read control means to repeat selected pop transactions and/or cause write control means to repeat selected push transactions.

Abstract: A global timebase system and method for a system-on-chip synchronizes multiple clock domains in each of a plurality of receiver modules by broadcasting a global timebase count value as Gray code over a global timebase bus. A global timebase generator includes a binary counter and a binary-to-Gray-code converter. Each receiver module registers the global timebase count value with its own local clock and includes a Gray-code-to-binary converter. The converted value, in binary form, may be used as least significant bits of a globally synchronized local timebase. Most significant bits may be generated by a local binary counter incremented at each 1-to-0 transition of the most significant bit of the global timebase count value.

Abstract: A direct memory access (DMA) engine schedules data transfer requests of a system-on-chip data processing system according to both an assigned transfer priority and the deadline for completing a transfer. Transfer priority is based on a hardness representing the penalty for missing a deadline. Priorities are also assigned to zero-deadline transfer requests in which there is a penalty no matter how early the transfer completes. If desired, transfer requests may be scheduled in timeslices according to priority in order to bound the latency of lower priority requests, with the highest priority hard real-time transfers wherein the penalty for missing a deadline is severe are given the largest timeslice. Service requests for preparing a next data transfer are posted while a current transaction is in progress for maximum efficiency. Current transfers may be preempted whenever a higher urgency request is received.

Abstract: A method and system for DMA transfer of data in scatter/gather mode. A table of buffer descriptors may be used to determine the next buffer to be used when a current buffer storing data that has been transferred or will be transferred and may be used in automatic buffer switching, which does not require processor intervention. Entries in the table of buffer descriptors are entered programmatically. The method and system also provide for hardware writing to table of packet descriptors which describes location and size of incoming data and can indicate whether a packet of data straddles two or more buffers, thus decoupling packet sizes from buffer sizes.

Abstract: A multiply accumulate unit (“MAC”) that performs operations on packed integer data. In one embodiment, the MAC receives 2 32-bit data words which, depending on the specified mode of operation, each contain either four 8-bit operands, two 16-bit operands, or one 32-bit operand. Depending on the mode of operation, the MAC performs either sixteen 8×8 operations, four 16×16 operations, or one 32×32 operation. Results may be individually retrieved from registers and the corresponding accumulator cleared after the read cycle. In addition, the accumulators may be globally initialized. Two results from the 8×8 operations may be packed into a single 32-bit register. The MAC may also shift and saturate the products as required.