Topic: asynchronous processing
Topic: concurrency
Topic: concurrency control by monitors
Topic: database transactions
Topic: managing shared memory
Topic: multiple activities in a user interface
Topic: multiple processors
Topic: parallel control statements
| |
Subtopic: replace-add
Quote: with replace-add, all processors can simultaneously modify a variable in constant time; requests combined in the network [»gottA4_1983]
| Quote: with concurrent loads, stores, and replace-adds, simultaneous loads may yield different results [»gottA4_1983]
| Quote: use replace-add for semaphores, readers/writers, parallel queues, and detecting completion [»gottA4_1983, OK]
| Quote: implement fetch-add with an omega network of shuffle switches [»gottA4_1983, OK]
| Subtopic: compare&swap, wait-free
Quote: concurrent data is wait-free if any process can complete in a finite number of steps; e.g., compare&swap but not test&set; reduce to consensus [»herlM1_1991]
| Subtopic: concurrent updates
Quote: for concurrent operations on B-trees, separate safe from unsafe nodes; an update to a safe node does not affect its ancestors [»bayeR_1977]
| Quote: algorithm for concurrent, tunable indexing of B*-trees; deadlock free [»bayeR_1977]
| Quote: time-split B-trees allow concurrent updates via short, independent atomic actions [»lomeD_1993]
| Subtopic: concurrent read
Quote: read methods of Guava monitors can execute concurrently
| Subtopic: parallel lookup
Quote: linear hashing is simpler than trees, much faster, easier locking for concurrency control [»litwW10_1980]
| Quote: constant-time lookup by parallel access, multi-level adaptive hashing; rehash when too many hash collisions [»browMH12_1992]
| Subtopic: parallel operations
Quote: parallel operations in CmLisp via a xector--a set of processors with one value per processor; e.g., xector add [»hillWD_1985]
| Quote: apply a xector of functions to all tuples with common xector indices; e.g., (.alpha.+ '{a->1 b->2} '{b-3 c->2}) => {b->5} [»hillWD_1985]
| Quote: in CmLisp, .beta. reduction applies a function to the values of a xector in logarithmic time; ignores indices [»hillWD_1985]
| Quote: set union, intersection, and universal qualifier can be unit-time operations on the Connection Machine [»hillWD_1985]
| Quote: the Connection Machine can shift arbitrarily large segments of data in unit time; with type codes can update pointers in unit time [»hillWD_1985]
| Quote: the Connection Machine can search for substrings in time proportional to the length of the substring [»hillWD_1985]
| Subtopic: efficient locks
Quote: ubiquitous synchronization via a space and time efficient meta-lock; 2 bits, 11 ops, no busy wait, flexible [»agesO11_1999]
| Quote: Guava avoids lock fields in object headers; only monitors require locks [»bacoDF10_2000]
| Subtopic: reference counting
Quote: multiprocessor, reference counting without synchronized operations; 100x reduction in maximal response time for garbage collection [»levaY10_2001]
| Subtopic: synchronization
Quote: firing squad synchronization problem--synchronize a line of processors to all enter the 'fire' state at the same time [»moorEF_1964]
| Quote: with immutability, consistency is a naming problem; consistency via domain relative addressing without serialization (more concurrency) [»walpJ6_1988]
|
Related Topics
Topic: asynchronous processing (30 items)
Topic: concurrency (33 items)
Topic: concurrency control by monitors (24 items)
Topic: database transactions (26 items)
Topic: managing shared memory (74 items)
Topic: multiple activities in a user interface (17 items)
Topic: multiple processors (10 items)
Topic: parallel control statements (12 items)
|