A Capabilities-Aware Framework for Using Computational Accelerators in Data-Intensive Computing

Multicore computational accelerators such as GPUs are now commodity components for highperformance computing at scale. While such accelerators have been studied in some detail as stand-alone computational engines, their integration in large-scale distributed systems raises new challenges and trade-offs. In this paper, we present an exploration of resource management alternatives for building asymmetric accelerator-based distributed systems. We present these alternatives in the context of a capabilities-aware framework for data-intensive computing, which uses an enhanced implementation of the MapReduce programming model for accelerator-based clusters, compared to the state of the art. The framework can transparently utilize heterogeneous accelerators for deriving high performance with low programming effort. Our work is the first to compare heterogeneous types of accelerators, GPUs and a Cell processors, in the same environment and the first to explore the trade-offs between compute-efficient and control-efficient accelerators on data-intensive systems. Our investigation shows that our framework scales well with the number of different compute nodes. Furthermore, it runs simultaneously on two different types of accelerators, successfully adapts to the resource capabilities, and performs 26.9% better on average than a static execution approach.

Characterisation of a C1qtnf5 Ser163Arg Knock-In Mouse Model of Late-Onset Retinal Macular Degeneration

A single founder mutation resulting in a Ser163Arg substitution in the C1QTNF5 gene product causes autosomal dominant late-onset retinal macular degeneration (L-ORMD) in humans, which has clinical and pathological features resembling age-related macular degeneration. We generated and characterised a mouse

Testing the effectiveness of commonly-used site curing regimes

A range of seven test methods was used to assess the effectiveness of curing on C30 and C50 Portland cement concretes. Curing was by formwork retention, wrapping in wet hessian or wrapping in polythene for periods of between one and seven days. Specimens from each mix were also subjected to both air and water storage.

The potential of electron beam radiation for simultaneous surface modification and bioresorption control of PLLA

Bioresorbable polymers have been widely investigated as materials exhibiting significant potential for successful application in the fields of tissue engineering and drug delivery. Further to the ability to control degradation, surface engineering of polymers has been highlighted as a key method central to their development. Previous work has demonstrated the ability of electron beam (e-beam) technology to control the degradation profiles and bioresorption of a number of commercially relevant bioresorbable polymers (poly-l-lactic acid (PLLA), Llactide/DL-lactide co-polymer (PLDL) and poly(lactic-co-glycolic acid (PLGA)). This work investigates the further potential of ebeam technology to impart added biofunctionality through the manipulation of polymer (PLLA) surface properties. PLLA samples were subjected to e-beam treatments in air, with varying beam energies and doses. Surface characterization was then performed using contact angle analysis, X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, and atomic force microscopy. Results demonstrated a significant increase in surface wettability post e-beam treatment. In correlation with this, XPS data showed the introduction of oxygen-containing functional groups to the surface of PLLA. Raman spectroscopy indicated chain scission in the near surface region of PLLA (as predicted). However, e-beam effects on surface properties were not shown to be dependent on beam energy or dose. E-beam irradiation did not seem to affect the surface roughness of PLLA as a direct consequence of the treatment.

Quotients, exactness, and nuclearity in the operator system category

We continue our study of tensor products in the operator system category. We define operator system quotients and exactness in this setting and refine the notion of nuclearity by studying operator systems that preserve various pairs of tensor products. One of our main goals is to relate these refinements of nuclearity to the Kirchberg conjecture. In particular, we prove that the Kirchberg conjecture is equivalent to the statement that every operator system that is (min,er)-nuclear is also (el,c)-nuclear. We show that operator system quotients are not always equal to the corresponding operator space quotients and then study exactness of various operator system tensor products for the operator system quotient. We prove that an operator system is exact for the min tensor product if and only if it is (min,el)-nuclear. We give many characterizations of operator systems that are (min,er)-nuclear, (el,c)-nuclear, (min,el)-nuclear and (el,max)-nuclear. These characterizations involve operator system analogues of various properties from the theory of C*-algebras and operator spaces, including the WEP and LLP.