2. Low Cost, High Performance, and Scalability: A New Approach to User-Level Distributed Shared Memory Patrick Anthony La Fratta WORTS 2005 15 December 2005 GUARANTEED! OR YOUR MONEY BACK!!!

3. Programming Models: Message-Passing

5. Programming Models: Shared Memory

6. Implementing a DSM System at the User Level

7. Implementing the DSM Client Initialization, Step 1: Get size of shared memory segment.

8. Initialization, Step 2: Map n pages into local memory. Implementing the DSM Client

9. Initialization, Step 3: Take away all access privileges from the shared segments. Implementing the DSM Client

10. Initialization, Step 4: Set up the segmentation fault handler. Implementing the DSM Client

11. Implementing the DSM System Application Reads Shared Address: Preview

12. Implementing the DSM System Shared address read, Step 1: Application reads shared address.

13. Implementing the DSM System Shared address read, Step 2: Control transferred to seg-fault handler.

14. Implementing the DSM System Shared address read, Step 3: Client contacts the server to get the page’s data.

15. Implementing the DSM System Shared address read, Step 4: Client grants read access privileges to application.

16. Implementing the DSM System Application Writes Shared Address: Preview

17. Implementing the DSM System Shared address write, Step 1: Application writes shared address.

18. Implementing the DSM System Shared address write, Step 2: Control transferred to seg-fault handler.

19. Implementing the DSM System Shared address write, Step 3: Client contacts server to with write notification.

20. Implementing the DSM System Shared address write, Step 4: Server calls back all other copies of pages being written.

21. Implementing the DSM System Shared address write, Step 5: Server indicates to client to proceed.

22. Implementing the DSM System Shared address write, Step 6: Client grants write privileges to application.

23. Implementing the DSM System Shared address write, Step 7: Later, the app detaches pages so others may use them.

24. Preliminary Results: All Pairs Shortest Paths Note: Results matched for all test cases, and all runs completed successfully.

25. System Profiling: All Pairs Shortest Paths

27. System Modifications and Extensions Better understanding of the trade-offs in the design of the interface. Efficient synchronization primitives through extended memory semantics with full/empty bits. Server-side per-page locking and client-side full- page flushing. * Speedups > 1! * System profiles resulted in:

28. Performance Results: Speedup for Various Configurations

29. Performance Results: Trends

30. Future Work Scalability: Enable clients to use more than one server. Peer-to-peer: Merge the server and client modules. Fault-tolerance: Checkpoint and Migration? Further testing: Implement and evaluate performance of other parallel applications. Questions?