Szymon OżógandGitHub ba118abfec improved caching for pointer arithmetics in ptx (#3922) ...

* improved caching for pointer arithmetics

* Add test for pointer arithmetics caching

* Refactor test

2024-04-04 07:33:48 -07:00

..

codegen

Tensor core ptx (#3894)

2024-04-04 07:32:31 -07:00

engine

graph schedule items (#4054)

2024-04-03 08:52:37 -07:00

features

search: fix counting of upcasts to ignore TC upcasts (#4045)

2024-04-02 19:52:05 -04:00

nn

Use exec_alu for lazy const folding (#4039)

2024-04-02 20:52:05 -04:00

renderer

improved caching for pointer arithmetics in ptx (#3922)

2024-04-04 07:33:48 -07:00

runtime

hotfix: make KFD timings right

2024-04-04 05:55:29 +00:00

shape

fix View.pad arg type (#3845)

2024-03-20 19:36:02 -04:00

__init__.py

move GlobalCounter to helpers (#4002)

2024-03-30 00:30:30 -04:00

buffer.py

move GlobalCounter to helpers (#4002)

2024-03-30 00:30:30 -04:00

device.py

use clang as default instead of llvm for cpu (#4035)

2024-04-02 00:02:18 -04:00

dtype.py

rename Scalar to ConstType and cast_scalar to as_const (#3946)

2024-03-26 22:39:58 -04:00

function.py

use *0+1 for Tensor.pow base case, remove function.Zero (#4023)

2024-03-31 19:20:44 -04:00

helpers.py

move GlobalCounter to helpers (#4002)

2024-03-30 00:30:30 -04:00

lazy.py

const fold ReduceOps (#4059)

2024-04-03 14:39:28 -04:00

ops.py

Ptx mulacc (#3937)

2024-04-04 00:15:25 -07:00

tensor.py

const fold cast unrealized_unpadded_const (#4047)

2024-04-03 12:31:24 -04:00