Ellen Spertus. Execution of Dataflow Programs on General-Purpose Hardware. Master 's thesis, Department of EECS, Massachusetts Institute of Technology,545 Tech. Square, Cambridge, MA, August 1992. To be expanded and released as Technical Report 1380. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
The J-Machine Multicomputer: An Architectural Evaluation - Noakes (1993) (36 citations) (Correct)
....general characteristics of the applications we studied. 4. 1 Programming systems Two J Machine programming systems were used to conduct the benchmark studies reported in this section: Tuned J and Concurrent Smalltalk (CST) Two other programming systems are also running on the machine: Id [12] [15] and PCN [16] Tuned J The J language serves as a system level programming language for the J Machine. It extends a pernode ANSI C environment with a small number of additional constructs for remote function invocation and synchronization. The user is responsible for data placement and task ....
SPERTUS, E. Execution of dataflow programs on generalpurpose hardware. MS Thesis, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, Aug. 1992.
Evaluating the Locality Benefits of Active Messages - Ellen Spertus And (1995) (3 citations) Self-citation (Spertus) (Correct)
....count is decremented. If the entry count has reached zero or if the thread is nonsynchronizing, a pointer to the thread is pushed on the LCV. The last item in the LCV is the address of the system code to swap in a new frame. An atomicity problem arises due to allowing inlets to interrupt threads [Spe92] Specifically, consider the case where the TAM fork t instruction is interrupted by an inlet that includes the instruction post t . If the thread and the inlet belong to the same codeblock activation, the synchronization counter for thread t may be improperly decremented. Similarly, the LCV ....
Ellen Spertus. Execution of Dataflow Programs on General-Purpose Hardware. Master 's thesis, Department of EECS, Massachusetts Institute of Technology,545 Tech. Square, Cambridge, MA, August 1992. To be expanded and released as Technical Report 1380.
Evaluation of Mechanisms for Fine-Grained Parallel Programs in .. - Ellen Spertus (1993) (19 citations) Self-citation (Spertus) (Correct)
....into two areas. One holds small arrays and activation frames. The other holds large arrays, which are spread across all the processors such that logically consecutive elements are on different processors. Memory is managed explicitly through library routines. The J Machine implementation of TAM [15] makes direct use of the hardware support for different priority levels. Threads run at the background priority level, allowing them to be quickly interrupted by messages arriving in the priority 0 queue. Priority 1 is currently not used. Because each priority level has its own register set, ....
E. Spertus. Execution of Dataflow Programs on GeneralPurpose Hardware. Master's thesis, Department of EECS, Massachusetts Institute of Technology, 545 Tech. Square, Cambridge, MA, August 1992. To be expandedand released as MIT AI Lab Technical Report 1380.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC