devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity

Vectorized the evaluation of tensor expression (using SSE, AVX, NEON, ...)

Browse Source 
Added the ability to parallelize the evaluation of a tensor expression over multiple cpu cores.
Added the ability to offload the evaluation of a tensor expression to a GPU.
This commit is contained in:
Benoit Steiner
2014-05-16 15:08:05 -07:00
parent 0320f7e3a7
commit 7402fea0a8
17 changed files with 719 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
unsupported/Eigen/CXX11/src/Tensor/TensorBase.h
View File

@@ -28,6 +28,7 @@ class TensorBase
typedef typename internal::traits<Derived>::Scalar Scalar;
typedef typename internal::traits<Derived>::Index Index;
typedef Scalar CoeffReturnType;
typedef typename internal::packet_traits<Scalar>::type PacketReturnType;
Derived& setZero() {
return setConstant(Scalar(0));
@@ -83,6 +84,17 @@ class TensorBase
return TensorCwiseBinaryOp<internal::scalar_sum_op<Scalar>, const Derived, const OtherDerived>(derived(), other.derived());
}
template<typename OtherDerived> EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE
const TensorCwiseBinaryOp<internal::scalar_difference_op<Scalar>, const Derived, const OtherDerived>
operator-(const OtherDerived& other) const {
return TensorCwiseBinaryOp<internal::scalar_difference_op<Scalar>, const Derived, const OtherDerived>(derived(), other.derived());
}
template <typename DeviceType>
TensorDevice<Derived, DeviceType> device(const DeviceType& device) {
return TensorDevice<Derived, DeviceType>(device, derived());
}
protected:
template <typename OtherDerived> friend class TensorBase;
EIGEN_DEVICE_FUNC