Files
pytorch/docs/source/notes/cuda.rst
Adam Paszke 4cc11066b2 Add torch.utils.data docs and improve notes (#460)
* Add torch.utils.data docs and improve notes
2017-01-17 14:51:05 -05:00

61 lines
2.1 KiB
ReStructuredText

CUDA semantics
==============
:mod:`torch.cuda` keeps track of currently selected GPU, and all CUDA tensors
you allocate will be created on it. The selected device can be changed with a
:any:`torch.cuda.device` context manager.
However, once a tensor is allocated, you can do operations on it irrespectively
of your selected device, and the results will be always placed in on the same
device as the tensor.
Cross-GPU operations are not allowed by default, with the only exception of
:meth:`~torch.Tensor.copy_`. Unless you enable peer-to-peer memory accesses
any attempts to launch ops on tensors spread across different devices will
raise an error.
Below you can find a small example showcasing this::
x = torch.cuda.FloatTensor(1)
# x.get_device() == 0
y = torch.FloatTensor(1).cuda()
# y.get_device() == 0
with torch.cuda.device(1):
# allocates a tensor on GPU 1
a = torch.cuda.FloatTensor(1)
# transfers a tensor from CPU to GPU 1
b = torch.FloatTensor(1).cuda()
# a.get_device() == b.get_device() == 1
z = x + y
# z.get_device() == 1
# even within a context, you can give a GPU id to the .cuda call
c = torch.randn(2).cuda(2)
# c.get_device() == 2
Best practices
--------------
Use pinned memory buffers
^^^^^^^^^^^^^^^^^^^^^^^^^
.. warning:
This is an advanced tip. You overuse of pinned memory can cause serious
problems if you'll be running low on RAM, and you should be aware that
pinning is often an expensive operation.
Host to GPU copies are much faster when they originate from pinned (page-locked)
memory. CPU tensors and storages expose a :meth:`~torch.Tensor.pin_memory`
method, that returns a copy of the object, with data put in a pinned region.
Also, once you pin a tensor or storage, you can use asynchronous GPU copies.
Just pass an additional ``async=True`` argument to a :meth:`~torch.Tensor.cuda`
call. This can be used to overlap data transfers with computation.
You can make the :class:`~torch.utils.data.DataLoader` return batches placed in
pinned memory by passing ``pinned=True`` to its constructor.