Skip to content

Phase 2 Experimental Plan

This plan defines the benchmark process used to evaluate multithreaded performance and records the completed execution setup.

Objective

Evaluate how execution time changes when the Java program computes the sum of cubes using 1, 2, 4, 6, and 8 threads.

Controlled Test Conditions

  • Use the same machine for all official runs.
  • Use the same input array size for all thread counts.
  • Ensure all array values are integers less than 100.
  • Measure only computation time (exclude setup and logging).

Execution Environment

  • Machine: MacBook Air
  • Processor: Apple M2
  • Cores: 8 total (4 performance + 4 efficiency)
  • RAM: 8 GB
  • Operating system: macOS Tahoe 26.3.1
  • Java runtime: OpenJDK 25 LTS (Temurin)

Measurement Method

  • Use System.nanoTime() before and after computation.
  • Convert elapsed time to milliseconds.
  • Execute each thread configuration repeatedly under the same setup.
  • The final recorded dataset contains 12 runs per configuration.
  • Compute average time for each configuration.

Metrics

  • Execution Time (ms)
  • Speedup = T(1) / T(n)
  • % Improvement = ((T(1) - T(n)) / T(1)) * 100

Analysis Questions

  • Where does performance improve the most?
  • At which thread count do gains flatten?
  • Does performance decrease at higher thread counts?
  • What overhead factors explain the trend (thread creation, scheduling, context switching, core limits)?

Observed Outcome

  • 2 threads showed overhead and performed worse than the baseline.
  • 4 threads provided only a modest improvement.
  • 6 and 8 threads produced the strongest performance gains.
  • 8 threads had the best average runtime overall, while 6 threads remained very close and slightly more stable in some runs.