Parallelism in High Performance Computing Victor Eijkhout CppCon 2024 ## I ntroduction This poster reports on ‘D2D’, a benchmark that explores elegance of expression and performance in the context context of a High Performance Computing ‘mini-application’. The same code has been implemented using a number of different approaches to parallelism. Implementations are discussed with performance results. ## arrays through 'mdspan', it is interesting to explore what C++ can offer for lower level performance critical operations. Scientific computing is an interesting test case since many algorithms are

## +24 ## High-Performance Numerical Integration in the Age of C++26 VINCENT REVERDY ## High-Performance Numerical Integration in the Age of C++26 Vincent Reverdy Laboratoire d'Annecy de Physique past, other languages do far better in terms of everything: functionality, ease of use, and even performance ## This talk The goal is NOT to revolutionize everything or show a library that beats everything methods ☑ Explicit adaptive Runge-Kutta schemes ☑ Slightly enhanced methods with interpolants ☑ High-accuracy interpolation inside integration steps ## I nterpolation with extra coefficients $ \beta_{i}

## Nim - the first high performance language with full support for hot codereloading at runtime by Viktor Kirilov ## Me, myself and I • my name is Viktor Kirilov - from Bulgaria • creator of doctest \begin{aligned}A ## I terators type CustomRange = object low: int high: int iterator items(range: CustomRange): int = var i = range.low while i <= range.high: yield i inc i iterator pairs(range: CustomRange): # uses CustomRange.items yield (i, char(i + ord('a')) for i, c in CustomRange(low: 1, high: 3): echo c # prints: b, c, d ## V ariants # This is an example how an abstract syntax tree

## +24 ## High-Performance Cross-Platform Architecture: C++20 Innovations ## NOAH STEIN ## About Me • 35-year career in video games and embedded software • Started using C++ in 1995 • First cross-platform moved into general-purpose registers for computations • Depending on the platform, may see a performance gain at this stage ## Quat Functions template

## +23 ## Symbolic Calculus for High-performance Computing From Scratch Using C++23 VINCENT REVERDY ## Symbolic Calculus for High-Performance Computing from Scratch using C++23 Vincent Reverdy Laboratoire know about optimization, performance, parallelism, ... ## What this talk is not about ■ Complicated maths (you are smart people, you can do it yourself) ☑ High-performance computing (you all know about ■ Symbolic calculus (derivatives, integrals) ☑ Full blown custom rule-based rewriting ## High-performance ■ Since formulas have the entire information on the mathematical AST, it's possible to

## Performance ## Lets dive into Performance issues • Everything in JavaScript defaults to being on the same thread. Too much work on main thread • Android nested layouts • Functions and objects defined

## PERFORMANCE MATTERS Emery Berger College of Information and Computer Sciences UMASS AMHERST (joint work with Charlie Curtsinger, Grinnell College) emeryberger.com, @emeryberger ## A short time ### un.bmp ## Performance used to be easy  Performance improvement in the '80s ## I I ## Performance improvement in [Image](/uploads/documents/6/9/a/5/69a5a7f2064c85b44eb3710c323581ae/p19_1.jpg) loading... ## Performance not easy anymore 

## Performance of Apache Ozone on NVMe Wei-Chiu Chuang (jojochuang) Ritesh Shukla (kerneltime) ## Agenda • Overview of how Ozone and how it scales • Why NVME is important for Ozone for scaling • Benefits Benefits of using NVME • Impala performance results from NVME clusters • Write path improvements results from NVME clusters • Summary • Questions ## Ozone Architecture |Slow startup based on size| |High Availability|Active - Active|Active - Standby| |Protocol Support|Hadoop / S3 API|Hadoop API| ## CLOUDERA

## How GitOps Boosts Business Performance: The Facts  ## I NTRODUCTION As cloud-native applications have become more prevalent competitive advantage with an increase in innovation. This positive effect is not limited to the performance of engineering teams. Technology, in particular cloud native technology like Kubernetes and its together six years of data drawn from over 31,000 technology professionals worldwide. It charts the performance of engineering teams across the world against four key measures: lead time for new features, failure

## 20 24 September 15 - 20 ## +24 ## Performance Engineering Being Friendly to Your Hardware ## I GNAS BAGDONAS ## Being Friendly to Your Hardware Performance Engineering A gentle introduction to hardware multilayer components ## Memory • What about HBM? Not 1024 bits, 8 x 128 instead. Bandwidth will only be high when there is sufficient stream of commands inflight • What about remote memory? CXL and vendor ble> From JESD 79-4 DDR4 specification Same capacity, different composition => different performance profile ## Memory • Memory system is in the uncore • Cores act as clients • Remote socket cores