You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: CHANGELOG.md
+20-2
Original file line number
Diff line number
Diff line change
@@ -1,12 +1,30 @@
1
-
### v3.3.20170222
1
+
### v3.5.20170718
2
+
- Feature parity with ArrayFire 3.5.
3
+
-`canny`: Canny Edge detector
4
+
-`Array.scalar`: Return the first element of the array
5
+
-`dot`: Now support option to return scalar
6
+
-`print_mem_info`: Prints memory being used / locked by arrayfire memory manager.
7
+
-`Array.allocated`: Returs the amount of memory allocated for the given buffer.
8
+
-`set_fft_plan_cache_size`: Sets the size of the fft plan cache.
9
+
10
+
- Bug Fixes:
11
+
-`sort_by_key` had key and value flipped in documentation.
12
+
13
+
- Improvements and bugfixes from upstream include:
14
+
- CUDA backend uses nvrtc instead of nvvm
15
+
- Performance improvements to arrayfire.reorder
16
+
- Faster unified backend
17
+
- You can find more information at arrayfire's [release notes](https://github.com/arrayfire/arrayfire/blob/v3.5.0/docs/pages/release_notes.md)
18
+
19
+
### v3.4.20170222
2
20
- Bugfix: Fixes typo in `approx1`.
3
21
- Bugfix: Fixes typo in `hamming_matcher` and `nearest_neighbour`.
4
22
- Bugfix: Added necessary copy and lock mechanisms in interop.py.
5
23
- Example / Benchmark: New conjugate gradient benchmark.
6
24
- Feature: Added support to create arrayfire arrays from numba.
7
25
- Behavior change: af.print() only prints full arrays for smaller sizes.
0 commit comments