# Any analog of torch.cuda.syncronize() in C++ API?

**URL:** <https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763>\
**Category:** C++\
**Created:** [February 9, 2019, 11:09am UTC](https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763 "2019-02-09T11:09:43Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Nikita\_Petrenko](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/n/ba8739/32.png) [@Nikita\_Petrenko](https://discuss.pytorch.org/u/Nikita_Petrenko)\
**Post date:** [February 9, 2019, 11:09am UTC](https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763/1 "2019-02-09T11:09:43Z")

</div>

I’ve been trying to record batch processing time, but stumbled upon asynchronous execution of torch operations. I know that I can sync with forward pass by calling loss.item() for example, but that looks ugly

Is there any analog of torch.cuda.syncronize() in C++ API?

---

<div class="post-metadata">

**Author:** ![albanD](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/alband/32/215_2.png) [@albanD](https://discuss.pytorch.org/u/albanD)\
**Post date:** [February 11, 2019, 10:12am UTC](https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763/2 "2019-02-11T10:12:16Z")

</div>

Hi,

You can call directly the `cudaDeviceSynchronize()` method as is done by the python code [here](https://github.com/pytorch/pytorch/blob/202eaa4ef40824d08ae88b3062b9c82d78794432/torch/csrc/cuda/Module.cpp#L212-L218).

---

<div class="post-metadata">

**Author:** ![Nikita\_Petrenko](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/n/ba8739/32.png) [@Nikita\_Petrenko](https://discuss.pytorch.org/u/Nikita_Petrenko)\
**Post date:** [February 11, 2019, 12:06pm UTC](https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763/3 "2019-02-11T12:06:25Z")

</div>

That’s nice, thank you!

Do you know if CPU code is also executed asynchronously?

---

<div class="post-metadata">

**Author:** ![albanD](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/alband/32/215_2.png) [@albanD](https://discuss.pytorch.org/u/albanD)\
**Post date:** [February 11, 2019, 12:40pm UTC](https://discuss.pytorch.org/t/any-analog-of-torch-cuda-syncronize-in-c-api/36763/4 "2019-02-11T12:40:46Z")

</div>

No all cpu code is synchronous !