HPTT: A High-Performance Tensor Transposition C++ Library
1–10 of 30 posts
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#2Why Ian this important? Not immediately clear to me.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#3Why Ian this important? Not immediately clear to me.
Faster tensor transposition, which can speed up applications that do manipulation and calculation with tensors. Tensors are typically seen in physics.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#4Why Ian this important? Not immediately clear to me.
Apparently TensorFlow uses Eigen (another library for Tensors) and HPTT executes up to 27.4 times faster than Eigen.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#5Why Ian this important? Not immediately clear to me.
Faster tensor transposition, which can speed up applications that do manipulation and calculation with tensors. Tensors are typically seen in physics.
For some reason, people use now the name "tensor" to refer to multidimensional arrays.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#6Why Ian this important? Not immediately clear to me.
Apparently TensorFlow uses Eigen (another library for Tensors) and HPTT executes up to 27.4 times faster than Eigen.
Note that they only compared their solution with Eigen’s tensor transposition functionality. HPTT is not a drop-in replacement.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#7Earlier quoted context omitted.
Faster tensor transposition, which can speed up applications that do manipulation and calculation with tensors. Tensors are typically seen in physics.
For some reason, people use now the name "tensor" to refer to multidimensional arrays.
In the case of physics, the name tensor is well deserved and used with purpose---the object are mathematically speaking tensors (and where called tensors before computers existed).
Their naming is unrelated to their physical representation in memory, even though they end up being arrays.
The same is true for various engineering disciplines.
For ML, it just turns out that they can talk about their objects and the corresponding operations also mathematically as tensors.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#8Earlier quoted context omitted.
Apparently TensorFlow uses Eigen (another library for Tensors) and HPTT executes up to 27.4 times faster than Eigen.
Note that they only compared their solution with Eigen’s tensor transposition functionality. HPTT is not a drop-in replacement.
Does anyone know how much time a typical TensorFlow model spends in transposition routines?
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#9Earlier quoted context omitted.
For some reason, people use now the name "tensor" to refer to multidimensional arrays.
In the case of physics, the name tensor is well deserved and used with purpose---the object are mathematically speaking tensors (and where called tensors before computers existed). Their naming is unrelated to their physical representation in memory, even though they end up being arrays. The same is true for various engineering disciplines. For ML, it just turns out that they can talk about their objects and the corr…
Except that none of these operations are related to the properties that make a tensor a tensor, as far as I know. People were already operating happily on multidimensional arrays before.
Re: HPTT: A High-Performance Tensor Transposition C++ Library
#10Earlier quoted context omitted.
Faster tensor transposition, which can speed up applications that do manipulation and calculation with tensors. Tensors are typically seen in physics.
For some reason, people use now the name "tensor" to refer to multidimensional arrays.
The reason probably is that tensors (in math/physics) can be represented as multi-indexed quantities, which are naturally expressed as multi-dimensional arrays in computers. 0-rank tensors are scalars, 1-rank tensors are vectors, 2-rank tensors can be represented as matrices (but strictly speaking are not matrices), and there are no specific names for higher-rank tensors.