“…On a distributed-memory architecture, such as the J-machine [19] or the CM-5 [52], we might instead maintain in software a fixed number of copies on different processors, which serve to spread out the load and reduce contention. Dally's ''distributed objects'' [18] or Chien's ''concurrent aggregates'' [14,13] might be useful substrates for implementing a multi-version memory on a distributed-memory machine.…”
Section: Implementation On Existing Architecturesmentioning
confidence: 99%
“…Our experiments were performed using Andrew Chien's Concurrent Aggregates (CA) language and multiprocessor simulator [14,13]. CA is an object-oriented language designed for fine-grained distributed-memory machines.…”
Section: The Simulatormentioning
confidence: 99%
“…The simulator also provides simple provisions to model resource contention (e.g., contention for the same memory module in the system) and network latency. 13 CA also provides synchronization primitives for implementing simple read/write locks for coherent shared memory and read/write pin operations for multi-version memory.…”
“…On a distributed-memory architecture, such as the J-machine [19] or the CM-5 [52], we might instead maintain in software a fixed number of copies on different processors, which serve to spread out the load and reduce contention. Dally's ''distributed objects'' [18] or Chien's ''concurrent aggregates'' [14,13] might be useful substrates for implementing a multi-version memory on a distributed-memory machine.…”
Section: Implementation On Existing Architecturesmentioning
confidence: 99%
“…Our experiments were performed using Andrew Chien's Concurrent Aggregates (CA) language and multiprocessor simulator [14,13]. CA is an object-oriented language designed for fine-grained distributed-memory machines.…”
Section: The Simulatormentioning
confidence: 99%
“…The simulator also provides simple provisions to model resource contention (e.g., contention for the same memory module in the system) and network latency. 13 CA also provides synchronization primitives for implementing simple read/write locks for coherent shared memory and read/write pin operations for multi-version memory.…”
“…I There can be a grave performance impact for the implicit synchronization this model imposes on threads even when they do not conflict. 4 While allowing for actors, pSather treats object-orientation and parallelism as orthogonal concepts, explicitly exposing Object orientation facilitates encapsulation of efficient parallel algorithms for machines with different characteristics. Compilers will not be able to perform the different algorithmic optimizations on the same source code necessary to achieve high performance on a single workstation, a T3D, and a network of workstations.…”
“…The group of actors have a common group name which may be used in conjunction with a more speci c parameter to access individual actors (or a subgroup of actors) within the group (cf. 15,12]). Given the name and a pattern, we get a particular member of the group represented by an actor address.…”
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.