Compilation Scheme Research Articles

Lexical Effect Handlers, Directly

Lexically scoping effect handlers is a language-design idea that equips algebraic effects with a modular semantics: it enables local-reasoning principles without giving up on the control-flow expressiveness that makes effect handlers powerful. However, we observe that existing implementations risk incurring costs akin to the run-time search for dynamically scoped handlers. This paper presents a compilation strategy for lexical effect handlers, adhering to the lexical scoping principle and targeting a language with low-level control over stack layout. Key aspects of this approach are formalized and proven correct. We embody the ideas in a language called Lexa: the Lexa compiler translates high-level effect handling to low-level stack switching. We evaluate the Lexa compiler on a set of benchmarks; the results suggest that it generates efficient code, reducing running-time complexity from quadratic to linear in some cases.

A Comparison Essay in the Context of the Management Phenomenon: Ancient Sumerian, Egyptian and Greek Civilizations

Abstract. The evolution of management ideas has been shaped by advancements spanning from ancient times to the present. Nevertheless, it is widely acknowledged that remnants of historical methods and practices persist in contemporary management approaches. The foundation of numerous management phenomena and theories in the contemporary era has been attempted to be elucidated by drawing upon ideas and practices that may be traced back to previous epochs, including ancient times. This study utilized secondary data and employed a compilation strategy. The objective of the study is to conduct a comparative analysis of the comprehension of management and society in the Sumerian, Egyptian, and Greek civilizations, which hold significant importance in ancient history. An attempt has been made to ascertain how the concept of management was developed in these three civilizations, the reasons that influenced management, and to identify and compare the key characteristics of management in each culture. The study aims to elucidate the disparities in governance and social structure among the three civilizations. The collected data have led to the conclusion that the geography of a civilization's settlement and the cultural and religious values of its society play a crucial role in shaping the system of government.

Spanish translation of the Expert Recommendations for Implementing Change (ERIC) compilation

BackgroundMost implementation science resources (e.g., taxonomies) are published in English. Linguistic inaccessibility creates a barrier to the conduct of implementation research among non-English-speaking populations, so translation of resources is needed. Translation into Spanish can facilitate widespread reach, given the large proportion of Spanish speakers around the world. This research aimed to systematically translate the Expert Recommendations for Implementation Change (ERIC) compilation into Spanish as an exemplar for the linguistic translation process.MethodsUsing the World Health Organization guidelines, this work translated the ERIC compilation strategy names, short definitions, and thematic clusters through a three-step process: 1) forward translation into Spanish by a native Spanish-speaking implementation scientist, 2) back-translation into English by a bilingual global health researcher, and 3) piloting via virtual focus group discussions with bilingual researchers not conducting implementation research. To achieve a generalizable translation, recruitment targeted a multicultural group of Spanish-speaking researchers. At the conclusion of each step, the transdisciplinary research team (N = 7) met to discuss discrepancies and refine translations. The Spanish version of the ERIC compilation was finalized through group consensus. Reflections from research team meetings and focus group discussions were synthesized qualitatively.ResultsGiven that dialectical nuances exist between Spanish-speaking regions, efforts prioritized universally accepted terminology. Team discussions focused on difficult translations, word choice, and clarity of concepts. Seven researchers participated in two focus groups, where discussion surrounded clarity of concepts, alternative word choice for Spanish translations, linguistic formality, grammar, and conciseness. Translation difficulties highlighted lack of precision in implementation science terminology, and the lack of conceptual clarity of words underscored limitations in the application of the compilation.ConclusionsThe work demonstrated the feasibility of translating implementation science resources. As one of the first systematic efforts to translate implementation resources, this study can serve as a model for additional efforts, including translation into other languages and the expansion to conceptual modifications. Further, this work yielded insights into the need to provide conceptual clarity in implementation science terminology. Importantly, the development of Spanish resources will increase access to conduct implementation research among Spanish-speaking populations.

A Function-Based Approach to Interactive High-Precision Volumetric Design and Fabrication

We present a novel function representation (F-Rep) based geometric modeling kernel tailor-made to support computer aided design (CAD) and fabrication of high resolution volumetric models containing hundreds of billions of voxel grid elements. Our modeling kernel addresses existing limitations associated with evaluating, storing, and accessing volumetric data produced by F-Reps in contexts outside of rendering. The result is an F-Rep modeling kernel well suited for CAD-based applications. Our kernel utilizes a sparse volume data structure to manage F-Rep data while efficient F-Rep evaluation is achieved through a combination of interval arithmetic (IA), just-in-time (JIT) compilation of user-defined functions, and massively parallel evaluation on the GPU. We employ IA as the basis for local pruning of the function evaluation tree to minimize total function evaluations, we use a novel JIT compilation scheme to optimize function execution, and we take advantage of GPU-parallelism to enhance computational throughput. We illustrate the kernel’s effectiveness in visualizing and slicing models with complex defining functions and detailed geometry, and utilize the geometry kernel to manufacture a physical part. Additionally, we present performance metrics across multiple hardware configurations demonstrating significant performance improvements over existing F-Rep geometry kernels, and we examine how our geometry kernel scales with computing power.

Quantum simulation of battery materials using ionic pseudopotentials

Ionic pseudopotentials are widely used in classical simulations of materials to model the effective potential due to the nucleus and the core electrons. Modeling fewer electrons explicitly results in a reduction in the number of plane waves needed to accurately represent the states of a system. In this work, we introduce a quantum algorithm that uses pseudopotentials to reduce the cost of simulating periodic materials on a quantum computer. We use a qubitization-based quantum phase estimation algorithm that employs a first-quantization representation of the Hamiltonian in a plane-wave basis. We address the challenge of incorporating the complexity of pseudopotentials into quantum simulations by developing highly-optimized compilation strategies for the qubitization of the Hamiltonian. This includes a linear combination of unitaries decomposition that leverages the form of separable pseudopotentials. Our strategies make use of quantum read-only memory subroutines as a more efficient alternative to quantum arithmetic. We estimate the computational cost of applying our algorithm to simulating lithium-excess cathode materials for batteries, where more accurate simulations are needed to inform strategies for gaining reversible access to the excess capacity they offer. We estimate the number of qubits and Toffoli gates required to perform sufficiently accurate simulations with our algorithm for three materials: lithium manganese oxide, lithium nickel-manganese oxide, and lithium manganese oxyfluoride. Our optimized compilation strategies result in a pseudopotential-based quantum algorithm with a total Toffoli cost four orders of magnitude lower than the previous state of the art for a fixed target accuracy.

Synthesis of and compilation with time-optimal multi-qubit gates

We develop a method to synthesize a class of entangling multi-qubit gates for a quantum computing platform with fixed Ising-type interaction with all-to-all connectivity. The only requirement on the flexibility of the interaction is that it can be switched on and off for individual qubits. Our method yields a time-optimal implementation of the multi-qubit gates. We numerically demonstrate that the total multi-qubit gate time scales approximately linear in the number of qubits. Using this gate synthesis as a subroutine, we provide compilation strategies for important use cases: (i) we show that any Clifford circuit on n qubits can be implemented using at most 2n multi-qubit gates without requiring ancilla qubits, (ii) we decompose the quantum Fourier transform in a similar fashion, (iii) we compile a simulation of molecular dynamics, and (iv) we propose a method for the compilation of diagonal unitaries with time-optimal multi-qubit gates, as a step towards general unitaries. As motivation, we provide a detailed discussion on a microwave controlled ion trap architecture with magnetic gradient induced coupling (MAGIC) for the generation of the Ising-type interactions.

Compilation and scaling strategies for a silicon quantum processor with sparse two-dimensional connectivity

Inspired by the challenge of scaling-up existing silicon quantum hardware, we propose a 2d spin-qubit architecture with low compilation overhead. The architecture is based on silicon nanowire split-gate transistors which form 1d chains of spin-qubits and allow the execution of two-qubit operations among neighbors. We introduce a silicon junction which can couple four nanowires into 2d arrangements via spin shuttling and Swap operations. We then propose a modular sparse 2d spin-qubit architecture with unit cells of diagonally-oriented squares with nanowires along the edges and junctions on the corners. Targeting noisy intermediate-scale quantum (NISQ) demonstrators, we show that the proposed architecture allows for compilation strategies which outperform methods for 1d chains, and exhibits favorable scaling properties which enable trading-off compilation overhead and colocation of control electronics within each square by adjusting the nanowire length. An appealing feature of the proposed architecture is its manufacturability using complementary-metal-oxide-semiconductor (CMOS) fabrication processes.

Optimizing DNN Compilation for Distributed Training With Joint OP and Tensor Fusion

This article proposes <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DisCo</i> , an automatic deep learning compilation module for data-parallel distributed training. Unlike most deep learning compilers that focus on training or inference on a single device, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DisCo</i> optimizes a DNN model for distributed training over multiple GPU machines. Existing single-device compilation strategies do not work well in distributed training, due mainly to communication inefficiency that they incur. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DisCo</i> generates optimized, joint computation operator and communication tensor fusion strategies to enable highly efficient distributed training. A GNN-based simulator is built to effectively estimate per-iteration training time achieved by operator/tensor fusion candidates. A backtracking search algorithm is driven by the simulator, navigating efficiently in the large strategy space to identify good operator/tensor fusion strategies that minimize distributed training time. We compare <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DisCo</i> with existing DL fusion schemes and show that it achieves good training speed-up close to the ideal, full computation-communication overlap case.

A surprisingly simple Lua compiler—Extended version

Dynamically-typed programming languages are often implemented using interpreters, which offer several advantages in terms of portability and flexibility of the implementation. However, as a language matures and its programs get bigger, programmers may seek compilers, to avoid the interpretation overhead. In this study, we present LuaAOT, a simple ahead-of-time compiler for Lua derived from the reference Lua interpreter. We describe two alternative compilation strategies. The first one exemplifies an old idea of using partial evaluation to produce a compiler based on an existing interpreter. Its contribution is to apply this idea to a well-established programming language. We show that with a quite modest effort it is possible to implement an efficient compiler that covers the entirety of Lua, including coroutines and tail calls. The whole implementation required less than 500 lines of new code. For this effort, we reduced the running time of our benchmarks from 20% to 60%. The second compilation strategy is based on function “outlining”, where each bytecode is implemented by a small subroutine. This strategy reduces executable sizes and compilation times, at the cost of not speeding up the running times as much as the first compilation method. • In this paper, we present a surprisingly simple ahead-of-time Lua compiler. • We describe a compilation strategy where jumps become gotos and another strategy for compilation targets that can’t use goto. • We also describe a compilation strategy that uses subroutines to reduce the generated code size.

Register allocation compilation technique for ASIP in 5G micro base stations

The currently available compilation techniques are for general computing and are not optimized for physical layer computing in 5G micro base stations. In such cases, the foreseeable data sizes and small code size are application specific opportunities for baseband algorithm optimizations. Therefore, the special attention can be paid, for example, the specific register allocation algorithm has not been studied so far. The compilation for kernel sub-routines of baseband in 5G micro base stations is our focusing point. For applications of known and fixed data size, we proposed a compilation scheme of parallel data accessing, while operands can be mainly allocated and stored in registers. Based on a small register group (48×32b), the target of our compilation scheme is the optimization of baseband algorithms based on 4×4 or smaller matrices, maximizing the utilization of register files, and eliminating the extra register data exchanging. Meanwhile, when data is allocated into register files, we used VLIW (Very Long Instruction Word) machine to hide the time of data accessing and minimize the cost of data accessing, thus the total execution time is minimum. Experiments indicate that for algorithms with small data size, the cost of data accessing and extra addressing can be minimized.

Expressivity of Planning with Horn Description Logic Ontologies

State constraints in AI Planning globally restrict the legal environment states. Standard planning languages make closed-domain and closed-world assumptions. Here we address open-world state constraints formalized by planning over a description logic (DL) ontology. Previously, this combination of DL and planning has been investigated for the light-weight DL DL-Lite. Here we propose a novel compilation scheme into standard PDDL with derived predicates, which applies to more expressive DLs and is based on the rewritability of DL queries into Datalog with stratified negation. We also provide a new rewritability result for the DL Horn-ALCHOIQ, which allows us to apply our compilation scheme to quite expressive ontologies. In contrast, we show that in the slight extension Horn-SROIQ no such compilation is possible unless the weak exponential hierarchy collapses. Finally, we show that our approach can outperform previous work on existing benchmarks for planning with DL ontologies, and is feasible on new benchmarks taking advantage of more expressive ontologies.

Compilation Forking: A Fast and Flexible Way of Generating Data for Compiler-Internal Machine Learning Tasks

Compiler optimization decisions are often based on hand-crafted heuristics centered around a few established benchmark suites. Alternatively, they can be learned from feature and performance data produced during compilation. However, data-driven compiler optimizations based on machine learning models require large sets of quality data for training in order to match or even outperform existing human-crafted heuristics. In static compilation setups, related work has addressed this problem with iterative compilation. However, a dynamic compiler may produce different data depending on dynamically-chosen compilation strategies, which aggravates the generation of comparable data. We propose compilation forking, a technique for generating consistent feature and performance data from arbitrary, dynamically-compiled programs. Different versions of program parts with the same profiling and compilation history are executed within single program runs to minimize noise stemming from dynamic compilation and the runtime environment. Our approach facilitates large-scale performance evaluations of compiler optimization decisions. Additionally, compilation forking supports creating domain-specific compilation strategies based on machine learning by providing the data for model training. We implemented compilation forking in the GraalVM compiler in a programming-language-agnostic way. To assess the quality of the generated data, we trained several machine learning models to replace compiler heuristics for loop-related optimizations. The trained models perform equally well to the highly-tuned compiler heuristics when comparing the geometric means of benchmark suite performances. Larger impacts on few single benchmarks range from speedups of 20% to slowdowns of 17%. The presented approach can be implemented in any dynamic compiler. We believe that it can help to analyze compilation decisions and leverage the use of machine learning into dynamic compilation.

Optimising the Stability in Plan Repair via Compilation

Plan repair is the problem of solving a given planning problem by using a solution plan of a similar problem. Plan repair problems can arise in execution contexts, that is, when an agent performing the plan has to deal with some unexpected contingency that makes the given plan invalid. Repairing a plan works often much better than replanning from scratch, and is crucial when plans have to be kept stable. There is no planning system until now that guarantees to find plans at the minimum distance from an input plan. This paper presents the first approach to such a problem; we indeed introduce a simple compilation scheme that converts a classical planning problem into another where optimal plans correspond to plans with the minimum distance from an input plan. Our experiments using a number of planners show that such a simple approach can solve the plan repair problem optimally and more effectively than replanning from scratch for a large number of cases. Last but not least, the approach proves competitive with LPG-ADAPT.

A high-performance compilation strategy for multiplexing quantum control architecture

Quantum computers have already shown significant potential to solve specific problems more efficiently than conventional supercomputers. A major challenge towards noisy intermediate-scale quantum computing is characterizing and reducing the various control costs. Quantum programming describes the process of quantum computation as a sequence, whose elements are selected from a finite set of universal quantum gates. Quantum compilation translates quantum programs to ordered pulses to the quantum control devices subsequently and quantum compilation optimization provides a high-level solution to reduce the control cost efficiently. Here, we propose a high-performance compilation strategy for multiplexing quantum control architecture. For representative benchmarks, the utilization efficiency of control devices increased by 49.44% on average in our work, with an acceptable circuit depth expansion executing on several real superconducting quantum computers of IBM.

Blockchain-Based Trust Verification and Streaming Service Awareness for Big Data-Driven 5G and Beyond Vehicle-to-Everything (V2X) Communication

Vigorous security and transparency with minimum latency and a high data rate are some of the key requirements for 5G and beyond vehicular networks especially in case of big data-driven vehicle-to-everything (V2X) communication applications which plays an important role in enabling advance intelligent transportation systems (ITS) and Internet of Vehicle (IoV) based network. The trace-ability in blockchain makes it one of the vital options, back-boned by the decentralized network to ensure independent security and privacy by exploiting the hash of a previous transaction with the time stamp and announcing it publicly for rejection of possible adversary attack. Therefore, in this work, we have opted for blockchains for trusted authentication and service awareness in 5G and beyond vehicle-to-everything (V2X) communication-based vehicular networks to stream blocks by certifying privacy and security. Here, we have considered blockchain as intermediary trusted agent to witness communication between a vehicle node and edge node, where edge node also uploads microservices along with calling figure of microservices to the blockchain. Furthermore, to ensure vehicle privacy with joint verification of vehicular network and server, we have devised a mechanism in which every vehicle nodes anonymously ask for services from the edge server along with signing an individual’s identity. Moreover, based on generated vehicle requests and service awareness, we devised an improved edge cache scheme for the prior compilation of services at each edge. In comparison to the previous state-of-art work, the result shows that our proposed work handles the requests more efficiently to achieve a better service request cache hit rate generated by vehicles at the edge server.

Compilation of Parallel Data Access for Vector Processor in Radio Base Stations

The data size and data feature under the cur-rent mainstream compilation schemes can be diversified, formal methods are thus mandatory. However, in a micro base station, as an embedded system, the typical feature or opportunity of baseband algorithms is its predictable small code and data sizes. The research is thus needed and there is so far no dedicated research on the related register alloca-tion algorithms. In this letter, we focus on compilation re-search for baseband kernel subroutines of 5G micro base stations. We proposed a parallel data access compilation scheme for applications with fixed and known data size, while operation data can be mostly stored and allocated in register file. The goal of the compilation scheme is to opti-mize algorithms based on 4x4 or smaller matrices, to reach the maximum register file utility, and eliminate the extra register data swapping. At the same time, we hide the data access time when allocating data into register file, and minimize the data access cost by using VLIW machine, so that the overall running time is minimum. Based on a small register group (48x32b) and the proposed compilation method, this letter implemented frequently used baseband subroutines (kernels) of 4x4 matrix with complex variables of 16b+16b and reached minimized run time. Results show that the extra addressing and data accessing cost is mini-mized for algorithms with small data size.

Compiling the Deeds of Salzburg Saints: The 12th-Century De episcopis Salisburgensibus and the Monastery of Admont

This chapter examines De episcopis Salisburgensibus, a late 12th-century collection of texts concerning the bishops and history of Salzburg that was seemingly compiled by monks belonging to the Benedictine monastery of Admont in Styria, Austria. It discusses the structure, sources, manuscript tradition and reception of this little-known collection, which focuses primarily on the ›glorious‹ early centuries of the archiepiscopal see but through the inclusion of the Lives of later bishops and a bishops’ list also stresses continuity in the sanctity and importance of the Salzburg institution. This latter message was particularly apt at the time of writing, which followed decades of upheaval due to conflict between the Salzburg archbishops and Emperor Frederick I. The compilation of this propagandistic collection by Admont monks was in keeping with the close historical and cultural relationship between the monastery and the archepiscopal see and the long-standing role of Admont as a centre for the composition and compilation of hagiographical and historiographical works concerning Salzburg and its bishops. Building upon a detailed appraisal of the manuscript evidence, this chapter shines light on the compilation strategies employed by the compilers of De episcopis Salisburgensibus and discusses the collection in the context of the special relationship between Salzburg and Admont.

Introduction. Medieval Biographical Collections in Comparison

This volume offers a comparative approach to collections of biographical texts from the Buddhist, Christian and Islamic worlds in the period 400-1600. It examines how both common and diverse compilation and literary strategies could be employed within and across different regions and religions. Throughout history, remarkable men and women stood out in a way that triggered an often-complex process of retelling their lives and preserving their memory through written narratives, and as a result they could become examples and sources of inspiration and identification for communities. This special issue looks at instances where such stories were preserved and presented within wider collections of biographies, where they could acquire new meanings and significance within an overarching narrative. The volume combines specialist case-studies on specific biographical collections from different regions with a number of co-written comparative essays that draw on and expand upon the conclusions of the individual studies. Within this comparative framework, the writing and compilation strategies of medieval compilers and the reception and audience of biographical collections are discussed at length. The particular potential of biographical collections to allow compilers across different cultures to evoke and promote particular ›visions of community‹ is teased out throughout the special issue.

Making weak memory models fair

Liveness properties, such as termination, of even the simplest shared-memory concurrent programs under sequential consistency typically require some fairness assumptions about the scheduler. Under weak memory models, we observe that the standard notions of thread fairness are insufficient, and an additional fairness property, which we call memory fairness, is needed. In this paper, we propose a uniform definition for memory fairness that can be integrated into any declarative memory model enforcing acyclicity of the union of the program order and the reads-from relation. For the well-known models, SC, x86-TSO, RA, and StrongCOH, that have equivalent operational and declarative presentations, we show that our declarative memory fairness condition is equivalent to an intuitive model-specific operational notion of memory fairness, which requires the memory system to fairly execute its internal propagation steps. Our fairness condition preserves the correctness of local transformations and the compilation scheme from RC11 to x86-TSO, and also enables the first formal proofs of termination of mutual exclusion lock implementations under declarative weak memory models.

Compilation of sparse array programming models

This paper shows how to compile sparse array programming languages. A sparse array programming language is an array programming language that supports element-wise application, reduction, and broadcasting of arbitrary functions over dense and sparse arrays with any fill value. Such a language has great expressive power and can express sparse and dense linear and tensor algebra, functions over images, exclusion and inclusion filters, and even graph algorithms. Our compiler strategy generalizes prior work in the literature on sparse tensor algebra compilation to support any function applied to sparse arrays, instead of only addition and multiplication. To achieve this, we generalize the notion of sparse iteration spaces beyond intersections and unions. These iteration spaces are automatically derived by considering how algebraic properties annotated onto functions interact with the fill values of the arrays. We then show how to compile these iteration spaces to efficient code. When compared with two widely-used Python sparse array packages, our evaluation shows that we generate built-in sparse array library features with a performance of 1.4× to 53.7× when measured against PyData/Sparse for user-defined functions and between 0.98× and 5.53× when measured against SciPy/Sparse for sparse array slicing. Our technique outperforms PyData/Sparse by 6.58× to 70.3×, and (where applicable) performs between 0.96× and 28.9× that of a dense NumPy implementation, on end-to-end sparse array applications. We also implement graph linear algebra kernels in our system with a performance of between 0.56× and 3.50× compared to that of the hand-optimized SuiteSparse:GraphBLAS library.