Neighborhoods Banding Together Reasoning Globally about Programs Lisa LippincottThe code here is written in a fantasy C++, with extensions supporting local reasoning.void foo() implementation { … … … bar();

1/66Bridging the Gap: Writing Portable Programs for CPU and GPU using CUDA Thomas Mejstrik Sebastian Woblistin 2/66Content 1 Motivation Audience etc.. Cuda crash course Quiz time 2 Patterns Oldschool tell me about afterwards7/66 Motivation Patterns The dark path Cuda proposal Thank you Why write programs for CPU and GPU Difference CPU/GPU Algorithms are designed differently Latency/Throughput Memory sense? Scope of the talk7/66 Motivation Patterns The dark path Cuda proposal Thank you Why write programs for CPU and GPU Difference CPU/GPU Why it makes sense? Library/Framework developers Embarrassingly

Rules • IPC stands for Intra-Process Communication • Programs stand for software running in a constrained environment ▪ CCI stands for: oContract oConcept oImplementationProblem Definition: Channel

Design patterns for error handling in C++ programs using parallel algorithms and executors Mark Hoemmen* mhoemmen@stellarscience.com CppCon 2020 * hoʊ’mən, or hœm’mən; he/himWho am I? • > 10 years

• Open for extension • Closed for modification • Implemented via delegated polymorphism Since programs that conform to the open–closed principle are changed by adding new code, rather than by changing changing existing code, they do not experience the cascade of changes exhibited by non-conforming programs. — Robert C. Martin (1996)Example: Drawing Shapes — The C Way Before OCP Shape Data Drawing Functions q.y() } -> std::same_as; { q.z() } -> std::same_as; };Supporting Concepts Arithmetic MutuallyArithmetic template concept Arithmetic = std::is_arithmetic_v;

14SYCL Implementations in Development Multiple Backends in Development There is development on supporting SYCL on even more low-level frameworks. For more information: http://sycl.tech Source Code IMG PowerVR Renesas R-Car NVIDIA GPUs Experimental XILINX FPGAs POCL (open-source OpenCL supporting CPUs and NVIDIA GPUs and more) Any CPU Experimental AMD GPUs NVIDIA GPUs Any CPU SYCL enables C++ Std, C, Python Libraries Libraries augment compiler optimizations for Performance Portable programs Use open standards to run Performance Portable code on new generation, or different vendor’s

data • P.11: Encapsulate messy constructs, rather than spreading through the code • P.12: Use supporting tools as appropriate • P.13: Use support libraries as appropriate Stroustrup - C++ safety -CppCon • A set of fundamental guarantees are standard • There are rules for composing code fragments supporting different guarantees • The set of guarantees assumed by and provided by a unit of code is stated upgrade billions of lines of code at once • We want to use several different profiles in larger programs • Mixing profiles is a generalization of the safe/unsafe oversimplification • “safe code” calling

Color { Green,Orange, Purple };Why are these categories important? • Actions • Allow input to programs that is unknown when the program was written • Performing an action has consequences • Affect how oven1 = Oven(oven_reserve_next_available()); oven1.TurnOn(); oven1.SetTemperature(375);Supporting new functions typedef int OVEN_HANDLE typedef int OVEN_ERR OVEN_HANDLE oven_reserve_next_available();

program to compare std::expected to exception handling. My requirements were: 1. Generate two C++ programs (except.cpp & result.cpp) that are isomorphic to each other. 2. Focus on propagation vs handling research on the topics of performance of exceptions and reliability of the resulting code in C++.✨ Supporting ✨ this work 232 Github sponsorship page For inquiry: estell.exceptions@gmail.comQuestions