Abstract:
A scheduling apparatus for dynamically setting a size of a rotating register of a local register file during runtime ids provided. The scheduling apparatus may include a determiner configured to determine whether a non-rotating register of a central register file is sufficient to schedule a program loop; a selector configured to select at least one local register file to which a needed non-rotating register is allocated in response to a determination that the non-rotating register of a central register file has a size which is sufficient to loop a program loop; a scheduler configured to schedule a non-rotating register of the at least one selected local register file.
Abstract:
An apparatus and method for calculating a physical address of a register in a processor are provided. The apparatus includes an offset calculator configured to calculate an offset between the physical address and a logical address of the register, based on a current iteration number and a size of a rotating register; an address calculator configured to calculate the physical address of the register by adding the calculated offset to the logical address of the register; and an address corrector configured to output a final physical address of the register based on the calculated physical address and the size of the rotating register.
Abstract:
Provided is a method of managing a register port, the method including performing scheduling on register ports that are used during a plurality of cycles to enable performing of a calculation; encoding data of the register ports according to results of the scheduling, the encoding of the data including, with respect to data of one of the register ports that does not have a schedule during one of the plurality of cycles, equally encoding the data of the one register port during the one cycle with data of an adjacent cycle of the one register port, the adjacent cycle being adjacent to the one cycle; and transmitting results of the encoding to a device that includes the register ports.
Abstract:
A method of measuring software performance includes inserting a performance measurement code into a source code, stalling a target system, on which the code is executed by a processor, and a performance counter based on the performance measurement code, transmitting performance data corresponding to a stalled time point when the target system and the performance counter are stalled to a host system configured to store the performance data corresponding to the stalled time point, and resuming execution of the source code by the target system and of the performance counter while the performance data is transmitted and stored.