News for hackers. Stuff that matters.

rcorcs

Thanks for the tips.

I have a patch of the LLVM Heat Printer being reviewed and will soon be available with LLVM.

Ask HN: Automatic Parallelization Haskell Core IR

I am interested in getting started with Haskell Core IR in order to implement this Automatic Parallelizer directly on its Core IR:http://homepages.dcc.ufmg.br/~fernando/publications/papers/Rocha18_SCP.pdfIs there a good introductory guide for working with algebraic simplification or rewriting rules on the Core IR?Thanks

( Read More... | 1 comment | 2 points | on HN )
Your comments are welcome!
Re:Show HN: Evaluating Computational Creativity on Tuesday October 18, 2016 @05:33PM (#12738932)
This repository implements the model proposed in the following paper:

http://dx.doi.org/10.13140/RG.2.2.31622.78403

Re:Show HN: My own C compiler (just for fun) on Tuesday June 14, 2016 @07:36PM (#11906239)
I know there are hundreds of C compilers and that I am just reinventing the wheel. But my main goal is just learning. I have already written another compiler before, but it was just for a toy language. For this reason, I would like to build a compiler for a "real-life" language, and C is an important but yet small language. I want to master the whole process of a fully working compiler for a "real-life" language, and afterwards continue to build on top of this knowledge, since I have been doing research on automatic parallelisation, and I am interested in optimising compilers in general.

Even if this project doesn't turn out to be useful for a lot of people, I hope at least to inspire a few people to tackle big problems.

Cycle detection in GraphX / Spark?

Does someone know how to efficiently implement this cycle detection algorithm in GraphX / Spark? http://dx.doi.org/10.13140/RG.2.1.1233.8640 https://en.wikipedia.org/wiki/Rocha-Thatte_cycle_detection_algorithm

( Read More... | no comments | 2 points | on HN )
I don't think the Stencil pattern is the best pattern for sorting algorithms, but I think it can be done.

The following link has a very simple example: https://github.com/pskel/pskel/wiki

For the stencil computation, if the input and output array together are greater than the GPU memory, there is an automatic and transparent method for tiling the data array into the GPU's memory, computing the stencil pattern in several chunks.
PSkel offers a high-level API for programming for heterogeneous parallel architectures, including both multi-core CPUs and GPUs.

PSkel is a high-performance framework for parallel skeletons. Using a high-level abstraction for parallel skeletons, PSkel releases the programmer from the responsibility of writing boiler-plate code for parallel programming in heterogeneous architectures, e.g., explicit synchronization and data exchanges between GPU memory and main memory. Furthermore, the framework translates the abstractions described using its application programming interface (API) into lowlevel C++ code compatible with Intel TBB, OpenMP and NVIDIA CUDA. PSkel's API is mainly based on a C++ template library that implements parallel skeletons and provides useful constructs for developing parallel applications. The framework provides an API for manipulating input and output data; specifying stencil computations; encapsulating memory management, computations, and runtime details.

Show HN: Parallel Programming Frameworks

I've been developing a few parallel and distributed frameworks.https://github.com/rcorcs/spp SPP: Structured Parallel Patterns is a framework based on algorithmic (parallel) skeletons. Currently it only supports multithreads, but I intend to keep working to give support to hybrid architectures, such as GPU, MIC, and MPI.https://github.com/rcorcs/hadoop-watershed (Hadoop) Watershed is a...

( Read More... | no comments | 3 points | on HN )

Show HN: Compiler development

I'm currently developing a compiler for Oberon (x86, x64, MIPS) and also a compiler for C.https://github.com/rcorcs/OberonChttps://github.com/rcorcs/rccContributions are welcome.

( Read More... | 1 comment | 4 points | on HN )
I have just updated to use Google Speech API v2.

Older »