For each project, I provide a number of questions and key points to be
answered/addressed in your reports. Use the given links as starting points.
Do some research and find more -good- references to deepen your knowledge
on the matter.

Remember:
  - Build a story
  - Narrow the focus and dig in one or two details
  - Give good examples / own experimentation
  - Show you learnt/understood the topic
  - Use your own words

---------------------------

Expression templates vs Smart expression templates:
  - What are expression templates (ET)? 
  - What are smart expression templates (SET)?
  - What is the basic idea behind them? (Give examples)
  - How do they improve over ET?
  - Which tools/libraries make use of them? (Give a general overview of these libraries)
  * Keywords: Expression Templates, smart expression templates, lazy evaluation, memory bound, compute bound.
  * Links:
     * Expression Templates by Todd Veldhuizen: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.43.248
     * Techniques for Scientific C++ by Todd Veldhuizen: www.cs.indiana.edu/pub/techreports/TR542.pdf
     * arxiv.org/pdf/1104.1729

---------------------------

Eigen / Armadillo
  - Introduction: What is it? Why was the project born? Target optimizations / use cases?
  - API
     + Operands and Datatypes 
     + Classes of operations provided
     + Features. Matrix properties. (triangularity, symmetry, etc)
     + (Good set of interesting) Examples
  - Technology behind it / which optimizations are applied?
  - Uses own implementation of computational kernels? Relies on libraries like BLAS?
  - Performance results. Comparison with other tools.
  * Keywords: Eigen/Armadillo, (expression) templates, BLAS.
  * Links: 
     * http://eigen.tuxfamily.org/index.php?title=Main_Page
     * http://arma.sourceforge.net/

---------------------------

Build-to-order BLAS (BTO)
  - Introduction, architecture and target operations
  - API
     - Operands
     - Operators and datatypes
     - Properties, storage, ...
  - Approach to loop fusion and tiling
  - Memory-accesses models, architecture representation
  - Experimental results, comparisons
  * Keywords: Build-to-order, BLAS, loop fusion, loop tiling
  * Links:
     * rintintin.colorado.edu/~karlini/pohll08.pdf

---------------------------

LLVM: 
  a) Intermediate Representation (IR) and optimizations (up to, not including instruction selection)
  b) Code generation and auto-vectorization

  - Introduction to LLVM.
  - General overview of the LLVM framework/architecture: from program to machine code.
  - Focus on the specific module(s) for each topic.
    a) Intermediate representation and the optimizations performed on this representation
    b) Code generation and autovectorization.
  * Keywords: LLVM, intermediate representation, vectorization, SIMD.
  * Links:
     * http://llvm.org/
     * http://llvm.org/docs/

---------------------------

Peephole optimization
  - General introduction to code generation and compiler optimizations.
  - What is the idea behind peephole optimization?
  - Give examples of characteristic peephole (local) optimizations.
  - Automatic generation of peephole optimizers.
  * Keywords: Code generation, peephole optimizers.
  * Links:
     * Compilers: Principles, Techniques, and Tools (Chapter 8). Available in the CS library.
     * Using Peephole Optimization on Intermediate Code: http://dspace.ubvu.vu.nl/bitstream/handle/1871/2606/11047.pdf
     * Automatic Generation of Peephole Superoptimizers: http://theory.stanford.edu/~aiken/publications/papers/asplos06.pdf

---------------------------

JIT
  - Concept
  - Use cases
  - Pros/cons
  - Examples: Java, llvm, web browsers, ...
  - Examples of optimizations and performance numbers
  * Keywords: just-in-time compilation, java virtual machine, hotspot, llvm, firefox
  * Links:
     * http://www.oracle.com/technetwork/java/whitepaper-135217.html (chapter 3)
     * http://llvm.org/docs/tutorial/LangImpl4.html

---------------------------

Roofline model: 
  a) memory-bound
  b)compute-bound operations

  - What is the roofline model?
  - Goals of the roofline model?
  - Which architecture information/knowledge is required?
  - How to get that information? (manufacturer, empirical, tools, ...).
  - Concept of arithmetic intensity, compute-bound vs memory-bound, ...
  - Give and idea and examples of at least 3 optimizations to improve 
     a) memory-bound
     b) compute-bound 
    code.
  * Keywords: roofline model, arithmetic intensity, memory- vs compute-bound, stream.
  * Links:
      * An Insightful Visual Performance Model for Multicore Architectures: http://www.eecs.berkeley.edu/~waterman/papers/roofline.pdf
      * http://www.spiral.net/software/roofline.html

---------------------------

ICC/GCC autovectorization:
  - Concept of vectorization and SIMD extensions
  - Compiler flags/pragmas related to the topic
  - Limitations
  - Keywords and coding techniques to help the compiler
  - Try out and illustrate with several examples.
  * Keywords: icc/gcc, compiler options, automatic vectorization, SIMD, aliasing.
  * Links:
     - Upload the white paper

---------------------------

OpenUH (OpenMP: under the hood)
  - Quick overview of basic OpenMP constructs: #pragma omp parallel for, scheduling, and variable scoping.
  - Basic ideas of pthreads (thread creation, execution, and passing arguments to the threads)  --> for presentation, not report )
  - How OpenUH translates the basic constructs to pthreads
  - Runtime functionality required in addition
  * Keywords: OpenMP, pthreads, shared/private clauses
  * Links:
     * pacman.cs.tsinghua.edu.cn/papers_cwg/openuh.pdf
     * Using OpenMP by B. Chapman et al. chapter 8

---------------------------

Nanos Mercurium compiler
  - General structure of the compiler
  - Support for trying out new ideas/extensions for OpenMP
  - Example of how to add support for a new OpenMP directive
  * Keywords: Nanos, Mercurium, OpenMP, runtime
  * Links:
    * personals.ac.upc.edu/aduran/papers/2004/mercurium_ewomp04.pdf
    * https://pm.bsc.es/