Wide vector execution in single thread mode for an out-of-order processor
A non-limiting example of a computer-implemented method for implementing wide vector execution for an out-of-order processor includes entering, by the out-of-order processor, a single thread mode. The method further includes partitioning, by the out-of-order processor, a vector register file into a plurality of register files, each of the plurality of register files being associated with a vector execution unit, the vector execution units forming a wide vector execution unit. The method further includes receiving, by a vector scalar register of the out-of-order processor, a wide vector instruction. The method further includes processing, by the wide vector execution unit, the wide vector instruction.
1. A computer-implemented method for wide vector execution for an out-of-order processor, the method comprising:
partitioning, by the out-of-order processor, a vector register file into a plurality of register files, each of the plurality of register files being associated with a vector execution unit, the vector execution units forming a wide vector execution unit, wherein each of the plurality of register files has three read wires and four write wires connected to its associated vector execution unit;
receiving, by a vector scalar register of the out-of-order processor, a wide vector instruction;
determining whether the out-of-order processor is operating a single thread mode or a multi-thread mode;
responsive to determining that the out-of-order processor is operating in the single thread mode, processing, by the wide vector execution unit, the wide vector instruction; and
responsive to determining that the out-of-order processor is operating in the multi-thread mode, preventing processing the wide vector instruction.
2. The computer-implemented method of claim 1 , wherein the multithread mode is a symmetric multiprocessing 2 (SMP2) mode.
3. The computer-implemented method of claim 1 , wherein the multithread mode is a symmetric multiprocessing 4 (SMP4) mode.
4. The computer-implemented method of claim 1 , wherein the wide vector instruction is received from a load-store unit of the out-of-order processor.
5. The computer-implemented method of claim 1 , wherein the wide vector instruction is larger than any one of the plurality of register files.
6. The computer-implemented method of claim 1 , wherein vector execution unit is 128 bits, and wherein the wide vector execution unit is 512 bits.
7. The computer-implemented method of claim 1 , wherein the vector register file is 128 bits and 128 entries, and wherein each of the plurality of register files is 128 bits and 32 entries.