Methods and apparatus to improve optimizing loops with predictable recurring memory reads

A non-transitory computer readable storage medium comprising first instructions which, when executed, cause a machine to at least:

• generate an optimization plan associated with a first loop and corresponding optimization parameters, the optimization plan including a set of one or more optimizations;
• identify the optimization plan as a candidate optimization plan when a quantity of PRMRs included in the first loop is greater than a threshold, the candidate optimization plan to convert one or more of the PRMRs into loop-invariant PRMRs, the candidate optimization plan to transform the first loop into a second loop, a single iteration of the second loop corresponding to a quantity of iterations of the first loop, the quantity of iterations based on a least common multiple of one or more lengths associated with the PRMRs;
• and
• determine the optimization parameters based on the candidate optimization plan;
• and
• generate second instructions to be executed by a processor, the second instructions based on processing the first loop with the set of one or more optimizations included in the candidate optimization plan.

A method to improve loop optimization with predictable recurring memory reads (PRMRs), comprising:

• generating an optimization plan associated with a first loop and corresponding optimization parameters, the optimization plan including a set of one or more optimizations;
• identifying the optimization plan as a candidate optimization plan when a quantity of PRMRs included in the first loop is greater than a threshold, the candidate optimization plan to convert one or more of the PRMRs into loop-invariant PRMRs, the candidate optimization plan to transform the first loop into a second loop, a single iteration of the second loop corresponding to a quantity of iterations of the first loop, the quantity of iterations based on a least common multiple of one or more lengths associated with the PRMRs;
• determining the optimization parameters based on the candidate optimization plan;
• and
• generating instructions to be executed by a processor, the instructions based on processing the first loop with the set of one or more optimizations included in the candidate optimization plan.