Efficiently exploring architectural design spaces via predictive modeling

Architects use cycle-by-cycle simulation to evaluate design choices and understand tradeoffs and interactions among design parameters. Efficiently exploring exponential-size design spaces with many interacting parameters remains an open problem: the sheer number of experiments renders detailed simulation intractable. We attack this problem via an automated approach that builds accurate, confident predictive design-space models. We simulate sampled points, using the results to teach our models the function describing relationships among design parameters. The models produce highly accurate performance estimates for other points in the space, can be queried to predict performance impacts of architectural changes, and are very fast compared to simulation, enabling efficient discovery of tradeoffs among parameters in different regions. We validate our approach via sensitivity studies on memory hierarchy and CPU design spaces: our models generally predict IPC with only 1-2% error and reduce required simulation by two orders of magnitude. We also show the efficacy of our technique for exploring chip multiprocessor (CMP) design spaces: when trained on a 1% sample drawn from a CMP design space with 250K points and up to 55x performance swings among different system configurations, our models predict performance with only 4-5% error on average. Our approach combines with techniques to reduce time per simulation, achieving net time savings of three-four orders of magnitude. Copyright © 2006 ACM.

Systematic Usage of Embedded Modelling Languages in Automated Model Transformation Chains

Annotation of programs using embedded Domain-Specific Languages (embedded DSLs), such as the program annotation facility for the Java programming language, is a well-known practice in computer science. In this paper we argue for and propose a specialized approach for the usage of embedded Domain-Specific Modelling Languages (embedded DSMLs) in Model-Driven Engineering (MDE) processes that in particular supports automated many-step model transformation chains. It can happen that information defined at some point, using an embedded DSML, is not required in the next immediate transformation step, but in a later one. We propose a new approach of model annotation enabling flexible many-step transformation chains. The approach utilizes a combination of embedded DSMLs, trace models and a megamodel. We demonstrate our approach based on an example MDE process and an industrial case study.

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.

Celtic languages in European Higher Education: CRAMLAP - Celtic, Regional and Minority Languages Abroad Project

The Celtic, Regional and Minority Languages Abroad
Project (CRAMLAP) is funded by the European Commission
to research the provision and pedagogy of regional
and minority languages outside their national borders in
Europe. The teaching of Celtic languages across Europe
was the focus in year one (2003-2004). This article summarizes
the qualitative data received in response to
questionnaires sent to institutions across Europe offering
Celtic Studies. Responses indicated that Celtic Studies
are quite widely available across Europe. The languages
are taught in comparative linguistics, linguistics and English
departments, with few dedicated Celtic departments
or sections outside the Celtic countries. Irish is supported
abroad by Irish government grant aid which will
become more widely available in the immediate future.
Many of the teachers have considerable experience, but
limited pedagogic training. The lack of suitable teaching
resources is the most commonly expressed concern.

Zipf and Type-Token rules for the English, Spanish, Irish and Latin languages

The Zipf curves of log of frequency against log of rank for a large English corpus of 500 million word tokens, 689,000 word types and for a large Spanish corpus of 16 million word tokens, 139,000 word types are shown to have the usual slope close to –1 for rank less than 5,000, but then for a higher rank they turn to give a slope close to –2. This is apparently mainly due to foreign words and place names. Other Zipf curves for highlyinflected Indo-European languages, Irish and ancient Latin, are also given. Because of the larger number of word types per lemma, they remain flatter than the English curve maintaining a slope of –1 until turning points of about ranks 30,000 for Irish and 10,000 for Latin. A formula which calculates the number of tokens given the number of types is derived in terms of the rank at the turning point, 5,000 for both English and Spanish, 30,000 for Irish and 10,000 for Latin.