Strategies for Energy-Efficient Resource Management of Hybrid Programming Models

Many scientific applications are programmed using hybrid programming models that use both message passing and shared memory, due to the increasing prevalence of large-scale systems with multicore, multisocket nodes. Previous work has shown that energy efficiency can be improved using software-controlled execution schemes that consider both the programming model and the power-aware execution capabilities of the system. However, such approaches have focused on identifying optimal resource utilization for one programming model, either shared memory or message passing, in isolation. The potential solution space, thus the challenge, increases substantially when optimizing hybrid models since the possible resource configurations increase exponentially. Nonetheless, with the accelerating adoption of hybrid programming models, we increasingly need improved energy efficiency in hybrid parallel applications on large-scale systems. In this work, we present new software-controlled execution schemes that consider the effects of dynamic concurrency throttling (DCT) and dynamic voltage and frequency scaling (DVFS) in the context of hybrid programming models. Specifically, we present predictive models and novel algorithms based on statistical analysis that anticipate application power and time requirements under different concurrency and frequency configurations. We apply our models and methods to the NPB MZ benchmarks and selected applications from the ASC Sequoia codes. Overall, we achieve substantial energy savings (8.74 percent on average and up to 13.8 percent) with some performance gain (up to 7.5 percent) or negligible performance loss.

Skin Dendritic Cell Targeting via Microneedle Arrays Laden with Antigen-Encapsulated Poly-D, L-lactide-co-Glycolide Nanoparticles Induces Efficient Antitumor and Antiviral Immune Responses

The efficacious delivery of antigens to antigen-presenting cells (APCs), in particular, to dendritic cells (DCs), and their subsequent activation remains a significant challenge in the development of effective vaccines. This study highlights the potential of dissolving microneedle (MN) arrays laden with nanoencapsulated antigen to increase vaccine immunogenicity by targeting antigen specifically to contiguous DC networks within the skin. Following in situ uptake, skin-resident DCs were able to deliver antigen-encapsulated poly-d,l-lactide-co-glycolide (PGLA) nanoparticles to cutaneous draining lymph nodes where they subsequently induced significant expansion of antigen-specific T cells. Moreover, we show that antigen-encapsulated nanoparticle vaccination via microneedles generated robust antigen-specific cellular immune responses in mice. This approach provided complete protection in vivo against both the development of antigen-expressing B16 melanoma tumors and a murine model of para-influenza, through the activation of antigen-specific cytotoxic CD8(+) T cells that resulted in efficient clearance of tumors and virus, respectively. In addition, we show promising findings that nanoencapsulation facilitates antigen retention into skin layers and provides antigen stability in microneedles. Therefore, the use of biodegradable polymeric nanoparticles for selective targeting of antigen to skin DC subsets through dissolvable MNs provides a promising technology for improved vaccination efficacy, compliance, and coverage.

Systematic Engineering of Uniform, Highly Efficient, Targeted and Shielded Viral-Mimetic Nanoparticles

In the past decades, numerous types of nanomedicines have been developed for the efficient and safe delivery of nucleic acid-based drugs for cancer therapy. Given that the destination sites for nucleic acid-based drugs are inside cancer cells, delivery systems need to be both targeted and shielded in order to overcome the extracellular and intracellular barriers. One of the major obstacles that has hindered the translation of nanotechnology-based gene-delivery systems into the clinic has been the complexity of the design and assembly processes, resulting in non-uniform nanocarriers with unpredictable surface properties and efficiencies. Consequently, no product has reached the clinic yet. In order to address this shortcoming, a multifunctional targeted biopolymer is genetically engineered in one step, eliminating the need for multiple chemical conjugations. Then, by systematic modulation of the ratios of the targeted recombinant vector to PEGylated peptides of different sizes, a library of targeted-shielded viral-mimetic nanoparticles (VMNs) with diverse surface properties are assembled. Through the use of physicochemical and biological assays, targeted-shielded VMNs with remarkably high transfection efficiencies (>95%) are screened. In addition, the batch-to-batch variability of the assembled targeted-shielded VMNs in terms of uniformity and efficiency is examined and, in both cases, the coefficient of variation is calculated to be below 20%, indicating a highly reproducible and uniform system. These results provide design parameters for engineering uniform, targeted-shielded VMNs with very high cell transfection rates that exhibit the important characteristics for in vivo translation. These design parameters and principles could be used to tailor-make and assemble targeted-shielded VMNs that could deliver any nucleic acid payload to any mammalian cell type.

An efficient unbounded lock-free queue for multi-core systems

The use of efficient synchronization mechanisms is crucial for implementing fine grained parallel programs on modern shared cache multi-core architectures. In this paper we study this problem by considering Single-Producer/Single- Consumer (SPSC) coordination using unbounded queues. A novel unbounded SPSC algorithm capable of reducing the row synchronization latency and speeding up Producer-Consumer coordination is presented. The algorithm has been extensively tested on a shared-cache multi-core platform and a sketch proof of correctness is presented. The queues proposed have been used as basic building blocks to implement the FastFlow parallel framework, which has been demonstrated to offer very good performance for fine-grain parallel applications. © 2012 Springer-Verlag.

Efficient Architectural Design Space Exploration via Predictive Modeling

Efficiently exploring exponential-size architectural design spaces with many interacting parameters remains an open problem: the sheer number of experiments required renders detailed simulation intractable.We attack this via an automated approach that builds accurate predictive models. We simulate sampled points, using results to teach our models the function describing relationships among design parameters. The models can be queried and are very fast, enabling efficient design tradeoff discovery. We validate our approach via two uniprocessor sensitivity studies, predicting IPC with only 1–2% error. In an experimental study using the approach, training on 1% of a 250-K-point CMP design space allows our models to predict performance with only 4–5% error. Our predictive modeling combines well with techniques that reduce the time taken by each simulation experiment, achieving net time savings of three-four orders of magnitude.