numba-dpex icon indicating copy to clipboard operation
numba-dpex copied to clipboard

Data Parallel Extension for Numba

Results 107 numba-dpex issues
Sort by recently updated
recently updated
newest added

The `dpnp.asarray` constructor should be supported inside a `dpjit` decorated function.

Currently, numba-dpex uses the same target context to generate kernels for all types of SPIR-V supporting devices. One of the limitations of the design is that there is no easy...

enhancement

`math.ceil` and `math.floor` [are supported within a numba kernel](https://intelpython.github.io/numba-dpex/latest/user_guides/kernel_programming_guide/supported-python-features.html#standard-library-modules) and when using it within a `CPython` interpreter it returns an `int`. But within a `dpex.kernel` it returns a `float64`, which...

user

The new dpjit decorator in numba-dpex should support the following dpnp array constructors: - [ ] `arange` - [ ] `linspace` - [x] `empty` - [x] `empty_like`: #928 - [...

enhancement
dpnp

Now functions from the сmath do not work on the kernel

The `kernel` function currently silently ignores Python exceptions and assertions. - [ ] The raise statement - [ ] The assert statement It is the correct behavior as a kernel...

enhancement

As reported in #756, some older generation GPUs do not support `float64` operations. In this case numba-dpex only raises a non-informative `SyclProgramCompilationError` exception without any specific message that makes it...

good first issue
user

The documentation currently states that [`numba_dpex` does not yet support SYCL private and constant memory](https://intelpython.github.io/numba-dpex/latest/user_guides/kernel_programming_guide/memory-management.html#private-and-constant-memory), but it does in fact support private memory (`dpex.private.array`).

documentation
user

Evaluate the [RNG feature](https://numba.pydata.org/numba-doc/latest/cuda/random.html) supported by `numba.cuda` and develop a plan on how to include support for similar functionality for `numba_dpex.kernel`. | Feature | numba.cuda| numba_dpex | | --- |...

enhancement

The execution scope of `group_barrier` needs to be determined based on the `group` argument. Currently it is hard-coded to `MemoryScope.WORK_GROUP` since numba-dpex only supports `group_barrier` on work groups. Once sub-groups...