Complete C++ Scientific Programming Bundle - 21 Hours!

Explore the struct versus class distinction in C++, where struct has public data by default and class has private; use struct for data containers and class for encapsulated objects.

Explore the C++11 initialization features, including initialize a list, initialize list constructors for containers, aggregate and uniform initialization, and auto type inference with for-each style vector traversal.

Learn how rvalue references enable move semantics and perfect forwarding to avoid costly copies, with move constructors, universal references, and efficient overloading.

Explore shared, weak, and unique pointers in C++: manage ownership, prevent memory leaks and cyclic references with weak pointers, use make_shared and custom deleters, and safely transfer or release ownership.

Manage resources in C++ using copy constructors, copy assignment operators, and the delete keyword. Construct in place with emplace back and use shared_ptr or unique_ptr for ownership and move semantics.

Explore the Crono library in C++11 for precision-neutral time and date handling, comparing system, steady, and high-resolution clocks, and mastering durations and time points.

Explore how operator new and operator delete allocate and free memory, overload them for dog and yellow dog, and the role of set new handler and virtual destructors.

Examine std::vector, deque, and list as sequence containers, and map and multimap as associative containers; learn iterators, algorithms, and performance tradeoffs like contiguous memory, cache locality, and splice operations.

Explore how the STL modifying algorithms change values and order, including copy, move, transform, swap, fill, remove, unique, reverse, rotate, permute, and shuffle.

Explore how reverse iterators work in the standard template library, convert to regular iterators with base, and print a vector in reverse, while noting iterator invalidation.

Explore removing elements from containers using the STL, focusing on vector and list techniques and erase-remove idiom. Understand iterator invalidation and removal for associative containers with algorithm-based and lambda-based removal.

Compare vector and deque in c++, detailing growth, reallocation, and usage rules: push front prefers deque; use vector for contiguous memory; consider data type, memory constraints, and pointer validity.

Learn to build graphs with the bushcraft adjacency list and attach vertex and edge properties. Use property maps to access these properties for algorithms like Dijkstra shortest path.

Learn how to package a factorial task into a packaged task, bind its parameter with a function object, and retrieve the result via a future in multi-threading.

Learn how OpenMP enables simple parallelization on multi-core systems, distinguish concurrent from true parallelism, and use a pragma to run code across multiple threads.

Learn how OpenMP handles data sharing before parallel regions, distinguishing shared and private data, and use firstprivate, lastprivate, default clauses and reduction to prevent race conditions and enable safe parallelization.

Discover how MPI enables distributing data across cluster servers to parallelize programs and speed up execution. Learn that MPI is a free, no-cost message passing interface API accessible to beginners.

Learn MPI simultaneous send and receive in C++, including buffer setup, data type, destination, tag, status, and a ring pattern, then introduce collective communication and broadcast.

Explore a rank-aware debug utility using variadic arguments and the underlying printf mechanism, and learn MPI collective operations like broadcast and barrier to synchronize all processes.

Overlay a two-dimensional Cartesian topology on an MPI communicator to map ranks to grid coordinates and use cart rank, cart coords, and cart shift for neighbor communication.

Explore how the gpu's parallel throughput powers data-parallel computations beyond cpu, and how kernels launch thousands of threads organized into grids, blocks, and threads on host-device systems via pci express.

Learn the kernel definition syntax with the global declaration to run on the device. Kernels return no value, use void, and threadIdx.x enables parallel indexing for operations like vector addition.

Learn to manage HPC software modules to set up software environments and run jobs on a system using module load, unload, and list commands.

Learn to write a PBS job script from a bash script using '#PBS' comments to request wall time, resources, and memory, then run your HPC commands.

Learn to write a Slurm job script named hybrid, requesting five minutes, four nodes, ten CPUs, one GB memory, loading Intel MPI and MPLX, and submit via sbatch.