Skip to content

Moinul laskar - Implemented optimized matrix multiplication#8

Open
priyoislam wants to merge 7 commits into
parallelcomputingabo:mainfrom
priyoislam:Moinul-Laskar
Open

Moinul laskar - Implemented optimized matrix multiplication#8
priyoislam wants to merge 7 commits into
parallelcomputingabo:mainfrom
priyoislam:Moinul-Laskar

Conversation

@priyoislam
Copy link
Copy Markdown

  • Both Blocked_matmul and Parallel_matmul were implemented straightforwardly from the pseudo-code
  • Separate test files for experimenting with different block sizes and threads
  • Challenges:
    • Blocked Multiplication gave inconsistent results with the same test case. Sometimes underperforming than naive multiplication, regardless of the block size used
    • Running the same test case consecutively resulted in highly variable outcomes. For example, the parallel version sometimes showed a speedup of 2–3x, but in the next run, it suddenly jumped to 8–10x.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant