#include <algorithm>
#include <cassert>
#include <vector>
#include <boost/iterator/iterator_facade.hpp>
#include "linalg/dynamicMatrix.hh"
#include "utilities/threading.hh"
Go to the source code of this file.
|
| class | Kaskade::NumaDenseMatrix< Entry > |
| | A dense matrix class tailored towards NUMA architectures. More...
|
| |
| class | Kaskade::NumaVector< Entry > |
| | A vector class tailored towards NUMA architectures. This vector distributes its entries in blocks of approximately equal size to the different nodes of NUMA machines in order to exploit the larger accumulated memory bandwidth for fast vector operations. This is most beneficial for large vectors exceeding the cache of the CPUs. Modern processors (2015-01-01) have up to 6MB 3rd level cache, so this is intended for vectors of more than 50,000 doubles, say. More...
|
| |
