The Manchester project has developed a powerful dataflow processor based on dynamic tagging. This processor is large enough to tackle realistic applications and exhibits impressive speedup for programs with sufficient parallelism.

No abstract

International audienceIn transactional memory, aborted transactions reduce performance, and waste computing resources. Ideally, concurrent execution of transactions should be optimally ordered to minimise aborts, but such an ordering is often either complex, or unfeasible, to obtain. This paper introduces a new technique called steal-on-abort, which aims to improve transaction ordering at runtime. Suppose transactions A and B conflict, and B is aborted. In general it is difficult to predict this first conflict, but once observed, it is logical not to execute the two transactions concurrently again. In steal-on-abort, the aborted transaction B is stolen by its opponent transaction A, and queued behind A to prevent concurrent execution of A and B. Without steal-on-abort, transaction B would typically have been restarted immediately, and possibly had a repeat conflict with transaction A. Steal-on-abort requires no application-specific information, modification, or offline pre-processing. In this paper, it is evaluated using a sorted linked list, red-black tree, STAMP-vacation, and Lee-TM. The evaluation reveals steal-on-abort is highly effective at eliminating repeat conflicts, which reduces the amount of computing resources wasted, and significantly improves performance

We present the results of an interpretation of 2D and 3D seismic from offshore Lebanon in which we identify a suite of 5 linearly distributed trails of fluid escape pipes with pockmarks at their upper terminus. These features transect the thick Messinian evaporites and root within prominent NE-SW oriented pre-salt folds. The pipe trails are oriented orthogonal to the strike of the pre-salt folds, with a synchronous initial expulsion episode in each trail dated at 1.7 Ma (±0.3 Ma), approximately coeval with the onset of salt-detached growth faulting along the basin margin. Each expulsion episode has been systematically offset to the NW away from the pre-salt fold by the flow of the salt, resulting in deformation of the fluid escape pipes in the salt. The orientation of the pipe trails thus provides a direct kinematic indicator for the flow direction of the salt layer during early stages of gravity-driven deformation of the salt and overburden, concomitant with basin margin uplift and tilting. The unidirectional NW oriented flow is recorded over a region of some 50 km width within the 2 translational domain of the salt tectonic deformation. Synchronicity in the onset of fluid expulsion from overpressured reservoirs within the pre-salt succession evidenced by these pipe trails and growth fault development at the basin margins implies that the pipe trails record the kinematics of the deforming salt layer throughout its post-Messinian phase of deformation. The deformed pipe trails demonstrate a Couette flow regime for the salt layer and document subtle changes in cumulative strain and velocity (2-4 mm/yr; ±0.3 mm/yr) over distances of a few km. It is proposed that this novel method of using fluid flow features as natural markers for the kinematics of deforming salt layers could be utilized in other parts of Eastern Mediterranean, as well as other salt basins on Earth.

While Transactional Memory (TM) research on sharedmemory chip multiprocessors has been flourishing over the last years, limited research has been conducted in the cluster domain. In this paper, we introduce a research platform for exploiting software TM on clusters. The Distributed Software Transactional Memory (DiSTM) system has been designed for easy prototyping of TM coherence protocols and it does not rely on a software or hardware implementation of distributed shared memory. Three TM coherence protocols have been implemented and evaluated with established TM benchmarks. The decentralized Transactional Coherence and Consistency protocol has been compared against two centralized protocols that utilize leases. Results indicate that depending on network congestion and amount of contention different protocols perform better.