River trail

Herhut, Stephan; Hudson, Richard L.; Shpeisman, Tatiana; Sreeram, Jaswanth

doi:10.1145/2509136.2509516

Cited by 26 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The hardware devices may be selected by the developer or left to be decided by the framework. The same approach is used by Ikra [22] for Ruby, River Trail [16] for JavaScript and ASDP [28] for ActionScript. Library parallelism is also implemented by overloading operators calling into specialized accelerator APIs.…”

Section: Related Workmentioning

confidence: 99%

Python programmers have GPUs too: automatic Python loop parallelization with staged dependence analysis

Jacob

Trinder

Singer

2019

Proceedings of the 15th ACM SIGPLAN International Symposium on Dynamic Languages

View full text Add to dashboard Cite

Python is a popular language for end-user software development in many application domains. End-users want to harness parallel compute resources effectively, by exploiting commodity manycore technology including GPUs. However, existing approaches to parallelism in Python are esoteric, and generally seem too complex for the typical end-user developer. We argue that implicit, or automatic, parallelization is the best way to deliver the benefits of manycore to end-users, since it avoids domain-specific languages, specialist libraries, complex annotations or restrictive language subsets. Autoparallelization fits the Python philosophy, provides effective performance, and is convenient for non-expert developers. Despite being a dynamic language, we show that Python is a suitable target for auto-parallelization. In an empirical study of 3000+ open-source Python notebooks, we demonstrate that typical loop behaviour 'in the wild' is amenable to auto-parallelization. We show that staging the dependence analysis is an effective way to maximize performance. We apply classical dependence analysis techniques, then leverage the Python runtime's rich introspection capabilities to resolve additional loop bounds and variable types in a just-intime manner. The parallel loop nest code is then converted to CUDA kernels for GPU execution. We achieve orders of magnitude speedup over baseline interpreted execution and some speedup (up to 50x, although not consistently) over CPU JIT-compiled execution, across 12 loop-intensive standard benchmarks. CCS Concepts • Software and its engineering → Dynamic compilers; Scripting languages; Parallel programming languages; • Computer systems organization → Heterogeneous (hybrid) systems.

show abstract

Section: Related Workmentioning

confidence: 99%

Python programmers have GPUs too: automatic Python loop parallelization with staged dependence analysis

Jacob

Trinder

Singer

2019

Proceedings of the 15th ACM SIGPLAN International Symposium on Dynamic Languages

View full text Add to dashboard Cite

show abstract

“…In contrast to Termite, SCP is statically typed, enabling advanced type-based optimizations. In non-process-oriented models, parallel closures [9] and RiverTrail [6] address important safety issues. SCP integrates a distributed, persistent data structure.…”

Section: Related Workmentioning

confidence: 99%

Distributed Programming via Safe Closure Passing

Haller

Miller

2016

Electron. Proc. Theor. Comput. Sci.

View full text Add to dashboard Cite

Programming systems incorporating aspects of functional programming, e.g., higher-order functions, are becoming increasingly popular for large-scale distributed programming. New frameworks such as Apache Spark leverage functional techniques to provide high-level, declarative APIs for in-memory data analytics, often outperforming traditional "big data" frameworks like Hadoop MapReduce. However, widely-used programming models remain rather ad-hoc; aspects such as implementation trade-offs, static typing, and semantics are not yet well-understood. We present a new asynchronous programming model that has at its core several principles facilitating functional processing of distributed data. The emphasis of our model is on simplicity, performance, and expressiveness. The primary means of communication is by passing functions (closures) to distributed, immutable data. To ensure safe and efficient distribution of closures, our model leverages both syntactic and typebased restrictions. We report on a prototype implementation in Scala. Finally, we present preliminary experimental results evaluating the performance impact of a static, type-based optimization of serialization.

show abstract

“…WebWorkers were introduced in HTML5 and help in parallel decoding by taking care of some task in the background and then passing the data to the main thread through message passing. Apart from WebWorkers, other technologies that aim to bring parallelism by passing the computation onto the GPU are : River Trail [Herhut et al 2013] and WebCL [KhronosGroup 2012] . River Trail provides a ParallelArray data structure with primitive operations that operate on it and offload the computation to the GPU and the underlying OpenCL implementation.…”

Section: Renderingmentioning

confidence: 99%

Animation on the web

Ahire

Evans

Blat

2015

Proceedings of the 20th International Conference on 3D Web Technology

View full text Add to dashboard Cite

The main motivation of this paper is to provide a current state and a brief overview of animation on the web. Computer animation is used in many fields and it has seen a lot of development in the recent years. With the widespread use of WebGL and the age of powerful modern hardware available on small devices, 3D rendering on the browser is now becoming commonplace. Computer Animation can be described as the rendering of objects on screen, which can change shape and properties with respect to time. There are many approaches to rendering animation on the web, but none of them yet provide a coherent approach in terms of transmission, compression and handling of the animation data on the client side (browser). And if computer animation has to become more accessible over the web, these challenges need to be addressed in the same "minimalistic manner (requirement wise)" as every other multimedia content has been addressed on the web. We aim to provide an overview of the current state of the art, while commenting on the shortcomings pertaining to current formats/approaches and discuss some of the upcoming standards and trends which can help with the current implementation.

show abstract

River trail

Cited by 26 publications

References 28 publications

Python programmers have GPUs too: automatic Python loop parallelization with staged dependence analysis

Python programmers have GPUs too: automatic Python loop parallelization with staged dependence analysis

Distributed Programming via Safe Closure Passing

Animation on the web

Contact Info

Product

Resources

About