v11.0.0b3
版本发布时间: 2022-05-26 15:35:12
cupy/cupy最新发布版本:v13.3.0(2024-08-22 15:42:45)
This is the release note of v11.0.0b3. See here for the complete list of solved issues and merged PRs.
We are running a Gitter chat for general discussions and quick questions. Feel free to join the channel to talk with developers and users!
Highlights
Support cuTensorNet as an einsum backend (#6677) (thanks @leofang!)
A new accelerator for CuPy has been added (CUPY_ACCELERATORS=cutensornet).
This feature requires cuquantum-python >= 22.03 and cuTENSOR >= 1.5.0. And is used to accelerate and support large array sizes in the cupy.linalg.einsum API.
Changes without compatibility
Drop Support for ROCm 4.2 (#6734)
CuPy v11 will drop support for ROCm 4.2. We recommend users to use ROCm 4.3 or 5.0 instead.
Drop Support for NumPy 1.18/1.19 and SciPy 1.4/1.5 (#6735)
As per NEP29, NumPy 1.18/1.9 support has been dropped on 2021. SciPy supported versions are the one released close to NumPy supported ones.
Changes
New Features
- Support cuTensorNet (from cuQuantum) as an
einsumbackend (#6677) - Add
cupy.poly(#6697) - Support cupy.ndarray subclassing - Part 1 - Direct constructor call (#6716)
Enhancements
- Support cuDNN 8.4 (#6641)
- Support cuTENSOR 1.5.0 (#6665)
- JIT: Use C++14 (#6670)
- Support cuTENSOR 1.5.0 (#6722)
- Drop support for ROCm 4.2 in CuPy v11 (#6734)
- Drop support for NumPy 1.18/1.19 and SciPy 1.4/1.5 in CuPy v11 (#6735)
- Fix compilation warning caused by
ifdef(#6739)
Performance Improvements
- Accelerate
bincount,histogram2d,histogramddwith CUB (#6701)
Bug Fixes
- Fix memory leak in the FFT plan cache during multi-threading (#6704)
- Fix
ifdeffor ROCm >= 4.2 (#6750)
Code Fixes
- JIT: Cosmetic change of
Dim3class (#6644)
Documentation
- Fix imports of
scatter_addexample (#6696) - Minor improvement on the array API docs (#6706)
- Document the returned benchmark object (#6712)
- Use exposed name in user guide (#6718)
Tests
- Xfail a test of
LOBPCGon ROCm 5.0+ (#6603) - CI: Update repo for libcudnn7 in cuda10.2 (#6708)
- Bump pinned mypy version (#6710)
- Follow
scipy==1.8.1sparse dot bugfix (#6727) - Support testing CUDA 11.6+ in FlexCI (#6731)
- Fix GPG key issue in FlexCI base image (#6738)
Contributors
The CuPy Team would like to thank all those who contributed to this release!
@asi1024 @Dahlia-Chehata @emcastillo @kmaehashi @leofang @takagi
1、 cupy_cuda102-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 61.63MB
2、 cupy_cuda102-11.0.0b3-cp310-cp310-manylinux2014_aarch64.whl 35.47MB
3、 cupy_cuda102-11.0.0b3-cp310-cp310-win_amd64.whl 43.28MB
4、 cupy_cuda102-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 60.08MB
5、 cupy_cuda102-11.0.0b3-cp37-cp37m-manylinux2014_aarch64.whl 33.73MB
6、 cupy_cuda102-11.0.0b3-cp37-cp37m-win_amd64.whl 43.31MB
7、 cupy_cuda102-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 63.32MB
8、 cupy_cuda102-11.0.0b3-cp38-cp38-manylinux2014_aarch64.whl 36.93MB
9、 cupy_cuda102-11.0.0b3-cp38-cp38-win_amd64.whl 43.41MB
10、 cupy_cuda102-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 61.56MB
11、 cupy_cuda102-11.0.0b3-cp39-cp39-manylinux2014_aarch64.whl 35.4MB
12、 cupy_cuda102-11.0.0b3-cp39-cp39-win_amd64.whl 43.4MB
13、 cupy_cuda110-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 76.46MB
14、 cupy_cuda110-11.0.0b3-cp310-cp310-win_amd64.whl 58.06MB
15、 cupy_cuda110-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 74.9MB
16、 cupy_cuda110-11.0.0b3-cp37-cp37m-win_amd64.whl 58.1MB
17、 cupy_cuda110-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 78.15MB
18、 cupy_cuda110-11.0.0b3-cp38-cp38-win_amd64.whl 58.19MB
19、 cupy_cuda110-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 76.39MB
20、 cupy_cuda110-11.0.0b3-cp39-cp39-win_amd64.whl 58.19MB
21、 cupy_cuda111-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 95.64MB
22、 cupy_cuda111-11.0.0b3-cp310-cp310-win_amd64.whl 78.21MB
23、 cupy_cuda111-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 94.08MB
24、 cupy_cuda111-11.0.0b3-cp37-cp37m-win_amd64.whl 78.25MB
25、 cupy_cuda111-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 97.33MB
26、 cupy_cuda111-11.0.0b3-cp38-cp38-win_amd64.whl 78.35MB
27、 cupy_cuda111-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 95.57MB
28、 cupy_cuda111-11.0.0b3-cp39-cp39-win_amd64.whl 78.34MB
29、 cupy_cuda112-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 76.88MB
30、 cupy_cuda112-11.0.0b3-cp310-cp310-win_amd64.whl 58.56MB
31、 cupy_cuda112-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 75.32MB
32、 cupy_cuda112-11.0.0b3-cp37-cp37m-win_amd64.whl 58.59MB
33、 cupy_cuda112-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 78.57MB
34、 cupy_cuda112-11.0.0b3-cp38-cp38-win_amd64.whl 58.69MB
35、 cupy_cuda112-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 76.81MB
36、 cupy_cuda112-11.0.0b3-cp39-cp39-win_amd64.whl 58.68MB
37、 cupy_cuda113-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 74.12MB
38、 cupy_cuda113-11.0.0b3-cp310-cp310-win_amd64.whl 55.35MB
39、 cupy_cuda113-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 72.56MB
40、 cupy_cuda113-11.0.0b3-cp37-cp37m-win_amd64.whl 55.39MB
41、 cupy_cuda113-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 75.81MB
42、 cupy_cuda113-11.0.0b3-cp38-cp38-win_amd64.whl 55.48MB
43、 cupy_cuda113-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 74.05MB
44、 cupy_cuda113-11.0.0b3-cp39-cp39-win_amd64.whl 55.47MB
45、 cupy_cuda114-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 82.39MB
46、 cupy_cuda114-11.0.0b3-cp310-cp310-win_amd64.whl 63.83MB
47、 cupy_cuda114-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 80.82MB
48、 cupy_cuda114-11.0.0b3-cp37-cp37m-win_amd64.whl 63.87MB
49、 cupy_cuda114-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 84.08MB
50、 cupy_cuda114-11.0.0b3-cp38-cp38-win_amd64.whl 63.96MB
51、 cupy_cuda114-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 82.32MB
52、 cupy_cuda114-11.0.0b3-cp39-cp39-win_amd64.whl 63.96MB
53、 cupy_cuda115-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 79.48MB
54、 cupy_cuda115-11.0.0b3-cp310-cp310-win_amd64.whl 60.88MB
55、 cupy_cuda115-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 77.91MB
56、 cupy_cuda115-11.0.0b3-cp37-cp37m-win_amd64.whl 60.92MB
57、 cupy_cuda115-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 81.17MB
58、 cupy_cuda115-11.0.0b3-cp38-cp38-win_amd64.whl 61.01MB
59、 cupy_cuda115-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 79.4MB
60、 cupy_cuda115-11.0.0b3-cp39-cp39-win_amd64.whl 61.01MB
61、 cupy_cuda116-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 79.52MB
62、 cupy_cuda116-11.0.0b3-cp310-cp310-win_amd64.whl 60.9MB
63、 cupy_cuda116-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 77.95MB
64、 cupy_cuda116-11.0.0b3-cp37-cp37m-win_amd64.whl 60.94MB
65、 cupy_cuda116-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 81.21MB
66、 cupy_cuda116-11.0.0b3-cp38-cp38-win_amd64.whl 61.03MB
67、 cupy_cuda116-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 79.44MB
68、 cupy_cuda116-11.0.0b3-cp39-cp39-win_amd64.whl 61.03MB
69、 cupy_rocm_4_3-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 36.34MB
70、 cupy_rocm_4_3-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 35MB
71、 cupy_rocm_4_3-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 37.83MB
72、 cupy_rocm_4_3-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 36.26MB
73、 cupy_rocm_5_0-11.0.0b3-cp310-cp310-manylinux1_x86_64.whl 54.41MB
74、 cupy_rocm_5_0-11.0.0b3-cp37-cp37m-manylinux1_x86_64.whl 53.07MB
75、 cupy_rocm_5_0-11.0.0b3-cp38-cp38-manylinux1_x86_64.whl 55.9MB
76、 cupy_rocm_5_0-11.0.0b3-cp39-cp39-manylinux1_x86_64.whl 54.33MB