Method for generating training data for cache memory design based on artificial intelligence and system using same

A computing device, the device comprising:

• a memory;
• and
• at least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory,
• wherein the at least one program comprises instructions for setting a reuse profile, setting a first selection reuse distance based on the reuse profile, setting a first load index and a first real reuse distance based on the first selection reuse distance, modifying the reuse profile according to setting the first real reuse distance, and setting a second selection reuse distance based on the modified reuse profile,
• wherein
• the reuse profile includes a plurality of reuse distances,
• each of the plurality of reuse distances represents a number of accesses to a non-duplicated address during two times of accesses of a duplicated address, which indicates a locality of access patterns to the cache memory for determining an optimized cache replacement policy how an old data is erased from the cache memory as the cache memory fetching a new data from a main memory,
• the setting the reuse profile comprises:
• setting a number of reuse distances and a minimum reuse distance;
• setting a first reuse distance and a second reuse distance based on a preset mathematical formula, the number of the reuse distances, and the minimum reuse distance;
• and
• setting a first probability for the first reuse distance and a second probability for the second reuse distance,
• the modifying the reuse profile comprises:
• subtracting a numerical value as many as a number of uses of the first real reuse distance from a first access number for the first real reuse distance included in the reuse profile;
• and
• modifying the first probability for the first real reuse distance based on the modified first access number,
• and
• the reuse profile is repeatedly modified by the at least one processor whenever the at least one processor sets a real reuse distance which includes, at least, the first real reuse distance.

A system for a cache memory design based on artificial intelligence, the system comprising:

• a training data generation unit for generating training data by using a reuse distance;
• and
• a cache structure design unit configured to design a cache structure for an application by using an artificial intelligence model trained based on the training data,
• wherein the training data generation unit comprises:
• a first module for generating the reuse profile;
• and
• a second module for setting a first selection reuse distance based on the reuse profile, setting a first load index and a first real reuse distance based on the first selection reuse distance, modifying the reuse profile according to setting the first real reuse distance, and setting a second selection reuse distance based on the modified reuse profile,
• wherein
• the reuse profile includes a plurality of reuse distances,
• each of the plurality of reuse distances represents a number of accesses to a non-duplicated address during two times of accesses of a duplicated address, which indicates a locality of access patterns to the cache memory for determining an optimized cache replacement policy how an old data is erased from the cache memory as the cache memory fetching a new data from a main memory,
• the setting the reuse profile comprises:
• setting a number of reuse distances and a minimum reuse distance;
• setting a first reuse distance and a second reuse distance based on a preset mathematical formula, the number of the reuse distances, and the minimum reuse distance;
• and
• setting a first probability for the first reuse distance and a second probability for the second reuse distance,
• the modifying the reuse profile comprises:
• subtracting a numerical value as many as a number of uses of the first real reuse distance from a first access number for the first real reuse distance included in the reuse profile;
• and
• modifying the first probability for the first real reuse distance based on the modified first access number,
• and
• the reuse profile is repeatedly modified by the at least one processor whenever the at least one processor sets a real reuse distance which includes, at least, the first real reuse distance.