Distributed Shared Memory

Readings

There will be no formal class this week. Instead, please read the following:

• Kai Li, Shared virtual memory on loosely coupled multiprocessors (1986).
  This thesis is the first major work on distributed shared memory. You don't need to follow all of the technical details in this thesis; instead, try to understand Li's framing of the DSM problem and the thread of his argument.

• Jelica Protic et al., "Distributed Shared Memory: Concepts and Systems" (1996)
  This paper reviews later work in the area of distributed shared memory. Read it to understand what issues dominated later research into DSM.

Questions

• What is the appeal of DSM?
• How is DSM related to virtual memory?
• How do concurrency issues affect the design of DSM systems?
• What sort of programs will benefit from DSM? What do not?
• What were the major problems addressed in DSM research following Kai Li's dissertation?
• Did these advances change the scope of environments and problems appropriate for DSM?
• Why aren't DSM systems commonly in use today?

Online Discussion

There is no lecture this week; however, there is a reading response and discussion!

• Please submit your response by Tuesday night (Sept. 23rd). You may submit by WebCT under "Response 2". Or, if you do not have access to WebCT, you may submit your assignment via email at soma at ccsl.carleton.ca, subject line "Response 2".

• We will have an online discussion of the readings through Friday, 5 PM, via the discussion for this wiki page. Everyone should plan to post at least one message. Ideally, you should all post at least one question and one response. Your group presentation grade will be based on your participation in this discussion.

Have fun this week, see you online and then Sept. 29th in class!

--Anil