|
About the hardware-backends category
|
|
0
|
954
|
January 22, 2021
|
|
Embrace tensor subclass as a Python device registration API
|
|
5
|
265
|
March 28, 2025
|
|
Does PyTorch support RTX 5090?
|
|
1
|
160
|
March 14, 2025
|
|
Custom kernels Intel XPU backend in LibTorch C++ API
|
|
13
|
210
|
March 2, 2025
|
|
OpenCL Backend - Important Updates
|
|
16
|
6782
|
February 15, 2025
|
|
Question of using DeviceGuard regarding the exclusive use of a device
|
|
0
|
53
|
December 20, 2024
|
|
OpenGL interoperability
|
|
1
|
256
|
December 19, 2024
|
|
[tac] Follow up: Inductor HW backend implementation
|
|
7
|
742
|
November 16, 2024
|
|
[RFC] Adding Triton Backend for Aten operators
|
|
0
|
404
|
November 4, 2024
|
|
DTensor random RNG state support for non-CUDA backends
|
|
2
|
135
|
October 18, 2024
|
|
How profiling Pytorch Using Nsight Compute?
|
|
2
|
361
|
October 16, 2024
|
|
Prior art on implementing a "print" op on hardware
|
|
0
|
39
|
October 3, 2024
|
|
RFC Proposal for CUDA-Accelerated Dynamic Time Warping (DTW) Implementation in PyTorch
|
|
2
|
398
|
September 20, 2024
|
|
Why so many HW backend and nobody cooperate?
|
|
6
|
439
|
September 17, 2024
|
|
OpenCL backend dev - questions/support
|
|
4
|
249
|
August 29, 2024
|
|
Using Nsight Systems to profile GPU workload
|
|
11
|
30270
|
August 22, 2024
|
|
Backend Fallbacks
|
|
1
|
3386
|
August 22, 2024
|
|
Intel GPU Enabling Status and Feature Plan
|
|
0
|
515
|
August 16, 2024
|
|
find_package(Torch REQUIRED) fails - 2.3.1 and nightly
|
|
10
|
638
|
August 6, 2024
|
|
ROCm vs OpenCL/dlprimitives
|
|
0
|
264
|
August 5, 2024
|
|
What are these `_foreach_` operators and why there is no fallback?
|
|
6
|
752
|
June 25, 2024
|
|
Overlapping device to host copy with GPU collectives
|
|
5
|
582
|
June 4, 2024
|
|
MPS working group?
|
|
2
|
363
|
March 12, 2024
|
|
Implementing OpenCL backend for pytorch
|
|
14
|
16036
|
March 1, 2024
|
|
Implementing _copy_from_and_resize for OpenCL backend
|
|
0
|
301
|
November 21, 2023
|
|
Possible to use custom backend to create TensorImpl that allows custom datatype?
|
|
0
|
335
|
October 25, 2023
|
|
Custom TensorImpl and TorchDynamo
|
|
1
|
537
|
September 10, 2023
|
|
Weight sharing on cuda
|
|
12
|
2002
|
August 31, 2023
|
|
Unable to port with isfinite(isxxx) series functions
|
|
1
|
472
|
July 31, 2023
|
|
How to support heterogeneous memories
|
|
0
|
514
|
June 29, 2023
|