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
* Implemented printing of `usm_ndarray`[#1013](https://github.com/IntelPython/dpctl/pull/1013), [#1043](https://github.com/IntelPython/dpctl/pull/1043), [#1060](https://github.com/IntelPython/dpctl/pull/1060)
22
+
* Implemented support for advanced indexing for `dpctl.tensor.usm_ndarray`[#1095](https://github.com/IntelPython/dpctl/pull/1095), [#1097](https://github.com/IntelPython/dpctl/pull/1097), [#1099](https://github.com/IntelPython/dpctl/pull/1099), [#1101](https://github.com/IntelPython/dpctl/pull/1101)
23
+
* Implemented support for platform listing in `dpctl.__main__` script [#1014](https://github.com/IntelPython/dpctl/pull/1014)
24
+
* Improved performance of `dpctl.tensor.asnumpy`[#1026](https://github.com/IntelPython/dpctl/pull/1026)
25
+
* Added `UsmNDArray_Make*` C-API for constructing `dpctl.tensor.usm_ndarray` from native allocations [#1050](https://github.com/IntelPython/dpctl/pull/1050), [#1067](https://github.com/IntelPython/dpctl/pull/1067)
26
+
* Added support for `dpctl.SyclDevice.native_vector_width_*` device descriptors [#1075](https://github.com/IntelPython/dpctl/pull/1075)
27
+
* Added ``dpctl::tensor::usm_ndarray::get_shape_vector`` and ``dpctl::tensor::usm_ndarray::get_strides_vector`` methods [#1090](https://github.com/IntelPython/dpctl/pull/1090)
28
+
12
29
13
30
### Changed
14
31
15
-
* Removed `dpctl.select_host_device`, `dpctl.has_host_device`, `dpctl.SyclDevice.is_host`, and `dpctl.SyclDevice.has_aspect_host` since support for host device has been removed in DPC++ 2023 and from SYCL 2020 spec [#1208](https://github.com/IntelPython/dpctl/pull/1028).
32
+
* Removed `dpctl.select_host_device`, `dpctl.has_host_device`, `dpctl.SyclDevice.is_host`, and `dpctl.SyclDevice.has_aspect_host` since support for host device has been removed in DPC++ 2023 and from SYCL 2020 spec [#1028](https://github.com/IntelPython/dpctl/pull/1028)
33
+
*`usm_ndarray`is made writable by default [#1012](https://github.com/IntelPython/dpctl/pull/1012), and writable flag is now checked by `__setitem__`.
34
+
* Added convenience signature for C++ utility function in "dpctl4pybind11.hpp" [#1016](https://github.com/IntelPython/dpctl/pull/1016)
35
+
* Improved error reported when attempting to submit kernel that uses a data-type unsupported by target device [#1018](https://github.com/IntelPython/dpctl/pull/1018), [#1040](https://github.com/IntelPython/dpctl/pull/1040)
36
+
* Updated C++ code to require DPC++ 2023.0.0 or newer [#1028](https://github.com/IntelPython/dpctl/pull/1028), [#1066](https://github.com/IntelPython/dpctl/pull/1066)
37
+
* The `dpctl.tensor.Device` class supports `print_device_info` method [#1029](https://github.com/IntelPython/dpctl/pull/1029), equality comparison, and hashing [#1048](https://github.com/IntelPython/dpctl/pull/1048)
38
+
* Updated version of pybind11 used to 2.10.2 [#1031](https://github.com/IntelPython/dpctl/pull/1031)
39
+
* Improved internal utility responsible for reduction of iteration space dimensionality [#1044](https://github.com/IntelPython/dpctl/pull/1044/), [#1054](https://github.com/IntelPython/dpctl/pull/1054)
40
+
* Changed return type of `DCPCTLUSM_GetPointerType` function in SyclInterface library [#1061](https://github.com/IntelPython/dpctl/pull/1061), [#065](https://github.com/IntelPython/dpctl/pull/1065)
41
+
* Updated supported version of DLPack to 0.8 [#1073](https://github.com/IntelPython/dpctl/pull/1073)
42
+
* Implemented queue cache per context/device pair and deployed it in `dpctl.memory`, `dpctl.tensor.from_dlpack` and `dpctl.tensor` array creation functions [#1076](https://github.com/IntelPython/dpctl/pull/1076), [#1079](https://github.com/IntelPython/dpctl/pull/1079)
* Fixed a typo in device-specific dtype mapping [#1015](https://github.com/IntelPython/dpctl/pull/1015)
54
+
* Fixed default device integer type to align with NumPy's behavior on Windows [#1017](https://github.com/IntelPython/dpctl/pull/1017)
55
+
* Fixed unexpected overflow in `dpctl.tensor.linspace` when one of the parameters is the largest floating point value [#1034](https://github.com/IntelPython/dpctl/pull/1034)
56
+
* Constructors `dpctl.tensor.empty`, `dpctl.tensor.zeros`, and `usm_ndarray` constructor itself no longer allow to create array with data-types not supported by targeted device [#1042](https://github.com/IntelPython/dpctl/pull/1042)
57
+
* Fixed parameter validation in `dpctl.SyclQueue` constructor [#1052](https://github.com/IntelPython/dpctl/pull/1052)
58
+
* Fixed `usm_type` of the resulting array in `dpctl.tensor.tril` and `dpctl.tensor.triu` functions [#1062](https://github.com/IntelPython/dpctl/pull/1062)
59
+
* Used DPC++ configuration files to ensure correct use of conda compiler toolchain on Linux [#1072](https://github.com/IntelPython/dpctl/pull/1072)
60
+
* Fixed issue with empty argument of `dpctl.tensor.meshgrid` function [#1080](https://github.com/IntelPython/dpctl/pull/1080/)
61
+
* Fixed linking problem on Windows enabling `dpctl` to be functional on Windows for devices not supporting some data types [#1083](https://github.com/IntelPython/dpctl/pull/1083)
0 commit comments