ADAM: A Decentralized Parallel Computer Architecture Featuring Fast Thread and Data Migration and a Uniform Hardware Abstraction

The furious pace of Moore's Law is driving computer architecture into a realm where the the speed of light is the dominant factor in system latencies. The number of clock cycles to span a chip are increasing, while the number of bits that can be accessed within a clock cycle is decreasing. Hence, it is becoming more difficult to hide latency. One alternative solution is to reduce latency by migrating threads and data, but the overhead of existing implementations has previously made migration an unserviceable solution so far. I present an architecture, implementation, and mechanisms that reduces the overhead of migration to the point where migration is a viable supplement to other latency hiding mechanisms, such as multithreading. The architecture is abstract, and presents programmers with a simple, uniform fine-grained multithreaded parallel programming model with implicit memory management. In other words, the spatial nature and implementation details (such as the number of processors) of a parallel machine are entirely hidden from the programmer. Compiler writers are encouraged to devise programming languages for the machine that guide a programmer to express their ideas in terms of objects, since objects exhibit an inherent physical locality of data and code. The machine implementation can then leverage this locality to automatically distribute data and threads across the physical machine by using a set of high performance migration mechanisms. An implementation of this architecture could migrate a null thread in 66 cycles -- over a factor of 1000 improvement over previous work. Performance also scales well; the time required to move a typical thread is only 4 to 5 times that of a null thread. Data migration performance is similar, and scales linearly with data block size. Since the performance of the migration mechanism is on par with that of an L2 cache, the implementation simulated in my work has no data caches and relies instead on multithreading and the migration mechanism to hide and reduce access latencies.

Bubble Behavior in a Taylor Vortex

We present an experimental study on the behavior of bubbles captured in a Taylor vortex. The gap between a rotating inner cylinder and a stationary outer cylinder is filled with a Newtonian mineral oil. Beyond a critical rotation speed (ω[subscript c]), Taylor vortices appear in this system. Small air bubbles are introduced into the gap through a needle connected to a syringe pump. These are then captured in the cores of the vortices (core bubble) and in the outflow regions along the inner cylinder (wall bubble). The flow field is measured with a two-dimensional particle imaging velocimetry (PIV) system. The motion of the bubbles is monitored by using a high speed video camera. It has been found that, if the core bubbles are all of the same size, a bubble ring forms at the center of the vortex such that bubbles are azimuthally uniformly distributed. There is a saturation number (N[subscript s]) of bubbles in the ring, such that the addition of one more bubble leads eventually to a coalescence and a subsequent complicated evolution. Ns increases with increasing rotation speed and decreasing bubble size. For bubbles of non-uniform size, small bubbles and large bubbles in nearly the same orbit can be observed to cross due to their different circulating speeds. The wall bubbles, however, do not become uniformly distributed, but instead form short bubble-chains which might eventually evolve into large bubbles. The motion of droplets and particles in a Taylor vortex was also investigated. As with bubbles, droplets and particles align into a ring structure at low rotation speeds, but the saturation number is much smaller. Moreover, at high rotation speeds, droplets and particles exhibit a characteristic periodic oscillation in the axial, radial and tangential directions due to their inertia. In addition, experiments with non-spherical particles show that they behave rather similarly. This study provides a better understanding of particulate behavior in vortex flow structures.

Model Order Reduction for Determining Bubble Parameters to Attain a Desired Fluid Surface Shape

In this paper, a new methodology for predicting fluid free surface shape using Model Order Reduction (MOR) is presented. Proper Orthogonal Decomposition combined with a linear interpolation procedure for its coefficient is applied to a problem involving bubble dynamics near to a free surface. A model is developed to accurately and efficiently capture the variation of the free surface shape with different bubble parameters. In addition, a systematic approach is developed within the MOR framework to find the best initial locations and pressures for a set of bubbles beneath the quiescent free surface such that the resultant free surface attained is close to a desired shape. Predictions of the free surface in two-dimensions and three-dimensions are presented.