Design and Evaluation of the Hamal Parallel Computer

Parallel shared-memory machines with hundreds or thousands of processor-memory nodes have been built; in the future we will see machines with millions or even billions of nodes. Associated with such large systems is a new set of design challenges. Many problems must be addressed by an architecture in order for it to be successful; of these, we focus on three in particular. First, a scalable memory system is required. Second, the network messaging protocol must be fault-tolerant. Third, the overheads of thread creation, thread management and synchronization must be extremely low. This thesis presents the complete system design for Hamal, a shared-memory architecture which addresses these concerns and is directly scalable to one million nodes. Virtual memory and distributed objects are implemented in a manner that requires neither inter-node synchronization nor the storage of globally coherent translations at each node. We develop a lightweight fault-tolerant messaging protocol that guarantees message delivery and idempotence across a discarding network. A number of hardware mechanisms provide efficient support for massive multithreading and fine-grained synchronization. Experiments are conducted in simulation, using a trace-driven network simulator to investigate the messaging protocol and a cycle-accurate simulator to evaluate the Hamal architecture. We determine implementation parameters for the messaging protocol which optimize performance. A discarding network is easier to design and can be clocked at a higher rate, and we find that with this protocol its performance can approach that of a non-discarding network. Our simulations of Hamal demonstrate the effectiveness of its thread management and synchronization primitives. In particular, we find register-based synchronization to be an extremely efficient mechanism which can be used to implement a software barrier with a latency of only 523 cycles on a 512 node machine.

Sparsely Faceted Arrays: A Mechanism Supporting Parallel Allocation, Communication, and Garbage Collection

Conventional parallel computer architectures do not provide support for non-uniformly distributed objects. In this thesis, I introduce sparsely faceted arrays (SFAs), a new low-level mechanism for naming regions of memory, or facets, on different processors in a distributed, shared memory parallel processing system. Sparsely faceted arrays address the disconnect between the global distributed arrays provided by conventional architectures (e.g. the Cray T3 series), and the requirements of high-level parallel programming methods that wish to use objects that are distributed over only a subset of processing elements. A sparsely faceted array names a virtual globally-distributed array, but actual facets are lazily allocated. By providing simple semantics and making efficient use of memory, SFAs enable efficient implementation of a variety of non-uniformly distributed data structures and related algorithms. I present example applications which use SFAs, and describe and evaluate simple hardware mechanisms for implementing SFAs. Keeping track of which nodes have allocated facets for a particular SFA is an important task that suggests the need for automatic memory management, including garbage collection. To address this need, I first argue that conventional tracing techniques such as mark/sweep and copying GC are inherently unscalable in parallel systems. I then present a parallel memory-management strategy, based on reference-counting, that is capable of garbage collecting sparsely faceted arrays. I also discuss opportunities for hardware support of this garbage collection strategy. I have implemented a high-level hardware/OS simulator featuring hardware support for sparsely faceted arrays and automatic garbage collection. I describe the simulator and outline a few of the numerous details associated with a "real" implementation of SFAs and SFA-aware garbage collection. Simulation results are used throughout this thesis in the evaluation of hardware support mechanisms.

Understanding the Mechanisms of Cell and Population Responses of Coccolithophores to Changing Ocean Chemistry

The calcifying coccolithophores have been proposed as a potentially vulnerable group in the face of increasing surface ocean CO2 levels. A full understanding of the likely responses of this group requires better mechanistic information on pH- and CO2-sensitive processes that underlie cell function at molecular, cellular and population levels. New findings on the mechanisms of pH homeostasis at a molecular and cellular level in both diatoms and coccolithophores are shaping our understanding of how these important groups may respond or acclimate to changing ocean pH. Critical parameters including intracellular pH homeostasis and cell surface pH will be considered. These studies are being carried out in parallel with genetic studies of natural oceanic populations to assess the natural genetic and physiological diversity that will underlie adaptation of populations in the long term.

Derivation of clinically compliant MSCs from CD105+, CD24-differentiated human ESCs

Adult tissue-derived mesenchymal stem cells ( MSCs) have demonstrated therapeutic efficacy in treating diseases or repairing damaged tissues through mechanisms thought to be mediated by either cell replacement or secretion of paracrine factors. Characterized, self- renewing human ESCs could potentially be an invariable source of consistently uniform MSCs for therapeutic applications. Here we describe a clinically relevant and reproducible manner of generating identical batches of hESC- derived MSC ( hESC- MSC) cultures that circumvents exposure to virus, mouse cells, or serum. Trypsinization and propagation of HuES9 or H1 hESCs in feeder- and serum-free selection media generated three polyclonal, karyotypically stable, and phenotypically MSC-like cultures that do not express pluripotency- associated markers but displayed MSC- like surface antigens and gene expression profile. They differentiate into adipocytes, osteocytes, and chondrocytes in vitro. Gene expression and fluorescence- activated cell sorter analysis identified CD105 and CD24 as highly expressed antigens on hESC- MSCs and hESCs, respectively. CD105+, CD24- monoclonal isolates have a typical MSC gene expression profiles and were identical to each other with a highly correlated gene expression profile ( r(2) >.90). We have developed a protocol to reproducibly generate clinically compliant and identical hESC- MSC cultures.

Comparative capacitative calcium entry mechanisms in canine pulmonary and renal arterial smooth muscle cells.

Experiments were performed to determine whether capacitative Ca(2+) entry (CCE) can be activated in canine pulmonary and renal arterial smooth muscle cells (ASMCs) and whether activation of CCE parallels the different functional structure of the sarcoplasmic reticulum (SR) in these two cell types. The cytosolic [Ca(2+)] was measured by imaging fura-2-loaded individual cells. Increases in the cytosolic [Ca(2+)] due to store depletion in pulmonary ASMCs required simultaneous depletion of both the inositol 1,4,5-trisphosphate (InsP(3))- and ryanodine (RY)-sensitive SR Ca(2+) stores. In contrast, the cytosolic [Ca(2+)] rises in renal ASMCs occurred when the SR stores were depleted through either the InsP(3) or RY pathways. The increase in the cytosolic [Ca(2+)] due to store depletion in both pulmonary and renal ASMCs was present in cells that were voltage clamped and was abolished when cells were perfused with a Ca(2+)-free bathing solution. Rapid quenching of the fura-2 signal by 100 microM Mn(2+) following SR store depletion indicated that extracellular Ca(2+) entry increased in both cell types and also verified that activation of CCE in pulmonary ASMCs required the simultaneous depletion of the InsP(3)- and RY-sensitive SR Ca(2+) stores, while CCE could be activated in renal ASMCs by the depletion of either of the InsP(3)- or RY-sensitive SR stores. Store depletion Ca(2+) entry in both pulmonary and renal ASMCs was strongly inhibited by Ni(2+) (0.1-10 mM), slightly inhibited by Cd(2+) (200-500 microM), but was not significantly affected by the voltage-gated Ca(2+) channel (VGCC) blocker nisoldipine (10 microM). The non-selective cation channel blocker Gd(3+) (100 microM) inhibited a portion of the Ca(2+) entry in 6 of 18 renal but not pulmonary ASMCs. These results provide evidence that SR Ca(2+) store depletion activates CCE in parallel with the organization of intracellular Ca(2+) stores in canine pulmonary and renal ASMCs.

Stiffener debonding mechanisms in post-buckled CFRP aerospace panels

This paper describes the fractographic analysis of five CFRP post-buckled skin/stringer panels that were tested to failure in compression. The detailed damage mechanisms for skin/stiffener detachment in an undamaged panel were characterised and related to the stress conditions during post-buckling; in particular the sites of peak twist (at buckling nodes) and peak bending moments (at buckling anti-nodes). The initial event was intralaminar splitting of the +45 degrees plies adjacent to the skin/stiffener interface, induced by high twist at a nodeline. This was followed by mode II delamination, parallel to +/- 45 degrees plies and then lengthwise (0 degrees) shear along the stiffener centreline. The presence of defects or damage was found to influence this failure process, leading to a reduction in strength. This research provides an insight into the processes that control post-buckled performance of stiffened panels and suggests that 2D models and element tests do not capture the true physics of skin/stiffener detachment: a full 3D approach is required.

Kinematic Analysis and Dimensional Synthesis of Exechon Parallel Kinematic Machine for Large Volume Machining

A parallel kinematic machine (PKM) topology can only give its best performance when its geometrical parameters are optimized. In this paper, dimensional synthesis of a newly developed PKM is presented for the first time. An optimization method is developed with the objective to maximize both workspace volume and global dexterity of the PKM. Results show that the method can effectively identify design parameter changes under different weighted objectives. The PKM with optimized dimensions has a large workspace to footprint ratio and a large well-conditioned workspace, hence justifies its suitability for large volume machining.

Parallel patterns + Macro Data Flow for multi-core programming

Data flow techniques have been around since the early '70s when they were used in compilers for sequential languages. Shortly after their introduction they were also consideredas a possible model for parallel computing, although the impact here was limited. Recently, however, data flow has been identified as a candidate for efficient implementation of various programming models on multi-core architectures. In most cases, however, the burden of determining data flow "macro" instructions is left to the programmer, while the compiler/run time system manages only the efficient scheduling of these instructions. We discuss a structured parallel programming approach supporting automatic compilation of programs to macro data flow and we show experimental results demonstrating the feasibility of the approach and the efficiency of the resulting "object" code on different classes of state-of-the-art multi-core architectures. The experimental results use different base mechanisms to implement the macro data flow run time support, from plain pthreads with condition variables to more modern and effective lock- and fence-free parallel frameworks. Experimental results comparing efficiency of the proposed approach with those achieved using other, more classical, parallel frameworks are also presented. © 2012 IEEE.

Conceptual Design and Workspace Analysis of an Exechon-inspired Parallel Kinematic Machine

Inspired by the commercial application of the Exechon machine, this paper proposed a novel parallel kinematic machine (PKM) named Exe-Variant. By exchanging the sequence of kinematic pairs in each limb of the Exechon machine, the Exe-Variant PKM claims an arrangement of 2UPR/1SPR topology and consists of two identical UPR limbs and one SPR limb. The inverse kinematics of the 2UPR/1SPR parallel mechanism was firstly analyzed based on which a conceptual design of the Exe-Variant was carried out. Then an algorithm of reachable workspace searching for the Exe-Variant and the Exchon was proposed. Finally, the workspaces of two example systems of the Exechon and the Exe-Variant with approximate dimensions were numerically simulated and compared. The comparison shows that the Exe-Variant possesses a competitive workspace with the Exechon machine, indicating it can be used as a promising reconfigurable module in a hybrid 5-DOF machine tool system.

Role of insulin and insulin-like peptides in bone formation: identification of bone-specific target genes and regulatory mechanisms, and characterization of the insulin-mimetic effect of vanadium

Tese de Doutoramento em Biologia, Especialidade em Biologia Molecular, Universidade do Algarve, 2008

Mass-wasting episodes in the geological evolution of the Azores islands : timing, recurrence, mechanisms and consequence

Tese de doutoramento (co-tutela), Geologia (Geodinâmica Interna), Faculdade de Ciências da Universidade de Lisboa, Faculté des Sciences D’Orsay-Université Paris-Sud, 2014

Parallel Coupled Micro-Macro Actuators

This thesis presents a new actuator system consisting of a micro-actuator and a macro-actuator coupled in parallel via a compliant transmission. The system is called the Parallel Coupled Micro-Macro Actuator, or PaCMMA. In this system, the micro-actuator is capable of high bandwidth force control due to its low mass and direct-drive connection to the output shaft. The compliant transmission of the macro-actuator reduces the impedance (stiffness) at the output shaft and increases the dynamic range of force. Performance improvement over single actuator systems was expected in force control, impedance control, force distortion and reduction of transient impact forces. A set of quantitative measures is proposed and the actuator system is evaluated against them: Force Control Bandwidth, Position Bandwidth, Dynamic Range, Impact Force, Impedance ("Backdriveability'"), Force Distortion and Force Performance Space. Several theoretical performance limits are derived from the saturation limits of the system. A control law is proposed and control system performance is compared to the theoretical limits. A prototype testbed was built using permanenent magnet motors and an experimental comparison was performed between this actuator concept and two single actuator systems. The following performance was observed: Force bandwidth of 56Hz, Torque Dynamic Range of 800:1, Peak Torque of 1040mNm, Minimum Torque of 1.3mNm. Peak Impact Force was reduced by an order of magnitude. Distortion at small amplitudes was reduced substantially. Backdriven impedance was reduced by 2-3 orders of magnitude. This actuator system shows promise for manipulator design as well as psychophysical tests of human performance.

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.

The neural processing of masked speech: Evidence for different mechanisms in the left and right temporal lobes

It has been previously demonstrated that extensive activation in the dorsolateral temporal lobes associated with masking a speech target with a speech masker, consistent with the hypothesis that competition for central auditory processes is an important factor in informational masking. Here, masking from speech and two additional maskers derived from the original speech were investigated. One of these is spectrally rotated speech, which is unintelligible and has a similar (inverted) spectrotemporal profile to speech. The authors also controlled for the possibility of “glimpsing” of the target signal during modulated masking sounds by using speech-modulated noise as a masker in a baseline condition. Functional imaging results reveal that masking speech with speech leads to bilateral superior temporal gyrus (STG) activation relative to a speech-in-noise baseline, while masking speech with spectrally rotated speech leads solely to right STG activation relative to the baseline. This result is discussed in terms of hemispheric asymmetries for speech perception, and interpreted as showing that masking effects can arise through two parallel neural systems, in the left and right temporal lobes. This has implications for the competition for resources caused by speech and rotated speech maskers, and may illuminate some of the mechanisms involved in informational masking.

The neural processing of masked speech: evidence for different mechanisms in the left and right temporal lobes

It has been previously demonstrated that extensive activation in the dorsolateral temporal lobes associated with masking a speech target with a speech masker, consistent with the hypothesis that competition for central auditory processes is an important factor in informational masking. Here, masking from speech and two additional maskers derived from the original speech were investigated. One of these is spectrally rotated speech, which is unintelligible and has a similar (inverted) spectrotemporal profile to speech. The authors also controlled for the possibility of "glimpsing" of the target signal during modulated masking sounds by using speech-modulated noise as a masker in a baseline condition. Functional imaging results reveal that masking speech with speech leads to bilateral superior temporal gyrus (STG) activation relative to a speech-in-noise baseline, while masking speech with spectrally rotated speech leads solely to right STG activation relative to the baseline. This result is discussed in terms of hemispheric asymmetries for speech perception, and interpreted as showing that masking effects can arise through two parallel neural systems, in the left and right temporal lobes. This has implications for the competition for resources caused by speech and rotated speech maskers, and may illuminate some of the mechanisms involved in informational masking.