Compiling Scientific Code Using Partial Evaluation

Scientists are faced with a dilemma: either they can write abstract programs that express their understanding of a problem, but which do not execute efficiently; or they can write programs that computers can execute efficiently, but which are difficult to write and difficult to understand. We have developed a compiler that uses partial evaluation and scheduling techniques to provide a solution to this dilemma.

Exploiting the Parallelism Exposed by Partial Evaluation

We describe an approach to parallel compilation that seeks to harness the vast amount of fine-grain parallelism that is exposed through partial evaluation of numerically-intensive scientific programs. We have constructed a compiler for the Supercomputer Toolkit parallel processor that uses partial evaluation to break down data abstractions and program structure, producing huge basic blocks that contain large amounts of fine-grain parallelism. We show that this fine-grain prarllelism can be effectively utilized even on coarse-grain parallel architectures by selectively grouping operations together so as to adjust the parallelism grain-size to match the inter-processor communication capabilities of the target architecture.

Partial Evaluation for Scientific Computing: The Supercomputer Toolkit Experience

We describe the key role played by partial evaluation in the Supercomputer Toolkit, a parallel computing system for scientific applications that effectively exploits the vast amount of parallelism exposed by partial evaluation. The Supercomputer Toolkit parallel processor and its associated partial evaluation-based compiler have been used extensively by scientists at M.I.T., and have made possible recent results in astrophysics showing that the motion of the planets in our solar system is chaotically unstable.

A Parallelizing Compiler Based on Partial Evaluation

We constructed a parallelizing compiler that utilizes partial evaluation to achieve efficient parallel object code from very high-level data independent source programs. On several important scientific applications, the compiler attains parallel performance equivalent to or better than the best observed results from the manual restructuring of code. This is the first attempt to capitalize on partial evaluation's ability to expose low-level parallelism. New static scheduling techniques are used to utilize the fine-grained parallelism of the computations. The compiler maps the computation graph resulting from partial evaluation onto the Supercomputer Toolkit, an eight VLIW processor parallel computer.

Partial oxidation of methane over the perovskite oxides

Ba0.5Sr0.5Co0.8Fe0.2O3-delta and Ba0.5Sr0.5Co0.8Ti0.2O3-delta oxides were synthesized by a combined EDTA-citrate complexing method. The catalytic behavior of these two oxides with the perovskite structure was studied during the reaction of methane oxidation. The pre-treatment with methane has different effect on the catalytic activities of both the oxides. The methane pre-treatment has not resulted in the change of the catalytic activity of BSCFO owing to its excellent reversibility of the perovskite structure resulting from the excellent synergistic interaction between Co and Fe in the oxide. However, the substitution with Ti on Fe-site in the lattice makes the methane pre-treatment have an obvious influence on the activity of the formed BSCTO oxide.

Gas phase catalytic partial oxidation of toluene in a microchannel reactor

Gas phase partial oxidation of toluene over V/Ti oxide catalysts has been successfully performed in a microchannel reactor, which provides very good mass and heat transfer conditions. With the elimination of hot spots, which are known as the most negative factors for partial oxidation of hydrocarbons, steady and uniform reaction conditions can be achieved in the catalyst bed by using, the microreactor. Since the best performance of the catalysts might be exploited, the selectivity of partial oxidation products of toluene has remarkably increased compared to the traditional packed fixed-bed reactor, even without the bother of modifying the catalysts, diluting the reactants or catalysts with inert contents to avoid hot spots or improve the diffusion and mixing. Furthermore, in virtue of its inherent safety features, when using pure oxygen as oxidant, the reactions were handled safety within the explosion limits in the microreactor. With TiO2 carried V2O5 as catalysts, the total selectivity of benzaldehyde and benzoic acid reaches around 60%, and the toluene conversion is about 10%. The conversion can go up without violent decline of selectivity, unlike most fixed bed reactors. Space time yield of 3.12 kg h(-1) L-1 calculated on the basis of the channel volume has been achieved. The influence of operating conditions has been investigated in detail in the microreactor. (c) 2005 Elsevier B.V. All rights reserved.

Mast Cell Curve-Response in Partial Achilles Tendon Rupture After 830 nm Phototherapy

Objective: the aim of this study was to quantify mast cells at different time intervals after partial Achilles tendon rupture in rats treated with low-level laser therapy (LLLT). Background data: There is a high incidence of lesions and ruptures in the Achilles tendon that can take weeks and even months to heal completely. As the mast cells help in the healing repair phase, and LLLT has favorable effects on this tissue repair process, study of this modality on the quantity of mastocytes in the ruptured tendon is relevant. Methods: Sixty Wistar rats were subjected to partial Achilles' tendon rupture by direct trauma, randomized into 10 groups, and then divided into the group treated with 80mW aluminum gallium arsenide infrared laser diode, continuous wave, 2.8W/cm(2) power density, 40J/cm(2) energy density, and 1.12J total energy, and the simulation group. Both the groups were subdivided according to the histological assessment period of the sample, either 6h, 12h, 24h, 2 days, or 3 days after the rupture, to quantify the mastocytes in the Achilles' tendon. Results: the group subjected to LLLT presented a greater quantity of mastocytes in the periods of 6h, 12h, 24h, 2 days, and 3 days after rupture, compared with the simulation groups, but differences were detected between the sample assessment periods only in the simulation group. Conclusions: LLLT was shown to increase the quantity of mastocytes in the assessment periods compared with the simulation groups.

Accelerated partial irradiation for breast cancer: Systematic review and meta-analysis of 8653 women in eight randomized trials

Background and purpose: Accelerated partial breast irradiation (APBI) is the strategy that allows adjuvant treatment delivery in a shorter period of time in smaller volumes. This study was undertaken to assess the effectiveness and outcomes of APBI in breast cancer compared with whole-breast irradiation (WBI). Material and methods: Systematic review and meta-analysis of randomized controlled trials of WBI versus APBI. Two authors independently selected and assessed the studies regarding eligibility criteria. Results: Eight studies were selected. A total of 8653 patients were randomly assigned for WBI versus APBI. Six studies reported local recurrence outcomes. Two studies were matched in 5 years and only one study for different time of follow-up. Meta-analysis of two trials assessing 1407 participants showed significant difference in the WBI versus APBI group regarding the 5-year local recurrence rate (HR = 4.54, 95% CI: 1.78-11.61, p = 0.002). Significant difference in favor of WBI for different follow-up times was also found. No differences in nodal recurrence, systemic recurrence, overall survival and mortality rates were observed. Conclusions: APBI is associated with higher local recurrence compared to WBI without compromising other clinical outcomes. (C) 2014 Elsevier Ireland Ltd. All rights reserved.

Agency and animation: the performance of interactive game characters

As an animator and practice-based researcher with a background in games development, I am interested in technological change in the video game medium, with a focus on the tools and technologies that drive game character animation and interactive story. In particular, I am concerned with the issue of ‘user agency’, or the ability of the end user to affect story development—a key quality of the gaming experience and essential to the aesthetics of gaming, which is defined in large measure by its interactive elements. In this paper I consider the unique qualities of the video game1 as an artistic medium and the impact that these qualities have on the production of animated virtual character performances. I discuss the somewhat oppositional nature of animated character performances found in games from recent years, which range from inactive to active—in other words, low to high agency. Where procedural techniques (based on coded rules of movement) are used to model dynamic character performances, the user has the ability to interactively affect characters in real-time within the larger sphere of the game. This game play creates a high degree of user agency. However, it lacks the aesthetic nuances of the more crafted sections of games: the short cut-scenes, or narrative interludes where entire acted performances are mapped onto game characters (often via performance capture)2 and constructed into relatively cinematic representations. While visually spectacular, cut-scenes involve minimal interactivity, so user agency is low. Contemporary games typically float between these two distinct methods of animation, from a focus on user agency and dynamically responsive animation to a focus on animated character performance in sections where the user is a passive participant. We tend to think of the majority of action in games as taking place via playable figures: an avatar or central character that represents a player. However, there is another realm of characters that also partake in actions ranging from significant to incidental: non-playable characters, or NPCs, which populate action sequences where game play takes place as well as cut scenes that unfold without much or any interaction on the part of the player. NPCs are the equivalent to supporting roles, bit characters, or extras in the world of cinema. Minor NPCs may simply be background characters or enemies to defeat, but many NPCs are crucial to the overall game story. It is my argument that, thus far, no game has successfully utilized the full potential of these characters to contribute toward development of interactive, high performance action. In particular, a type of NPC that I have identified as ‘pivotal’3—those constituting the supporting cast of a video game—are essential to the telling of a game story, particularly in genres that focus on story and characters: adventure games, action games, and role-playing games. A game story can be defined as the entirety of the narrative, told through non-interactive cut-scenes as well a interactive sections of play, and development of more complex stories in games clearly impacts the animation of NPCs. I argue that NPCs in games must be capable of acting with emotion throughout a game—in the cutscenes, which are tightly controlled, but also in sections of game play, where player agency can potentially alter the story in real-time. When the animated performance of NPCs and user agency are not continuous throughout the game, the implication is that game stories may be primarily told through short movies within games, making it more difficult to define video games animation as a distinct artistic medium.