# Numba 0.54.0 RC2 and llvmlite v0.37.0 RC2 **URL:** \ **Category:** Announcements\ **Created:** [July 19, 2021, 11:03am UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810 "2021-07-19T11:03:13Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![esc](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/esc/32/49_2.png) [@esc](https://numba.discourse.group/u/esc)\ **Post date:** [July 19, 2021, 11:03am UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/1 "2021-07-19T11:03:14Z")
Dear all, on behalf of the Numba community I am super excited to announce the immediate availability of the second round of release candidates: Numba 0.54.0 RC2 and llvmlite 0.37.0 RC2! In addition to the availability of wheels from PyPi (now also available for aarch64/RaspberryPi) the NumPy version number is now clamped at `1.20.*` for stability purposes and the lineinfo emission for debugging purposes has been fixed. For details, please refer to the change-logs at: [Numba 0.54.0 RC2 release notes.](https://numba.readthedocs.io/en/0.54.0rc2/release-notes.html) and [llvmlite v0.37.0 RC2 release notes](https://llvmlite.readthedocs.io/en/v0.37.0rc2/release-notes.html). Conda packages can be installed from the `numba` channel on [anaconda.org](http://anaconda.org/) with `conda` : `conda install -c numba numba=0.54.0rc2` And `conda install -c numba llvmlite=v0.37.0rc2` Binary wheels are available from PyPi as pre-releases. You should be able to install them with: `pip install --pre numba` And `pip install --pre llvmlite` If you have time and inclination, please do test these release candidates with your own libraries and tools and report back any issues you may encounter! Especially folks with a RaspberryPi are encouraged to try the wheels! Thank you again to all contributors for helping to improve Numba! ---
**Author:** ![gmarkall](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/gmarkall/32/42_2.png) [@gmarkall](https://numba.discourse.group/u/gmarkall)\ **Post date:** [July 21, 2021, 1:28pm UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/2 "2021-07-21T13:28:49Z")
I’ve tested the wheels on a Jetson AGX Xavier (Aaarch64 + Volta GPU) and get no unexpected fails (just some CUDA fails that we know about on the Jetson already, [CUDA: Skip IPC tests on ARM by gmarkall · Pull Request #7189 · numba/numba · GitHub](https://github.com/numba/numba/pull/7189) and [CUDA: Fix test\_pinned on Jetson by gmarkall · Pull Request #7190 · numba/numba · GitHub](https://github.com/numba/numba/pull/7190)), and the installation worked as expected. So - AArch64 wheels are looking good to me. ---
**Author:** ![gmarkall](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/gmarkall/32/42_2.png) [@gmarkall](https://numba.discourse.group/u/gmarkall)\ **Post date:** [July 22, 2021, 4:16pm UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/3 "2021-07-22T16:16:39Z")
I can confirm that the conda-packged 0.54RC2 is good with most RAPIDS libraries (cuDF, cuML, cuGraph, cuSpatial) - I’m yet to test with cuSignal, but will be checking that next. ---
**Author:** ![gmarkall](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/gmarkall/32/42_2.png) [@gmarkall](https://numba.discourse.group/u/gmarkall)\ **Post date:** [July 23, 2021, 11:12am UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/4 "2021-07-23T11:12:44Z")
cuSignal also seems to be fine with Numba 0.54RC2. ---
**Author:** ![brunomagacho](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/brunomagacho/32/342_2.png) [@brunomagacho](https://numba.discourse.group/u/brunomagacho)\ **Post date:** [July 23, 2021, 11:21pm UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/5 "2021-07-23T23:21:34Z")
It has some integration with dask also? ---
**Author:** ![esc](https://yyz2.discourse-cdn.com/free1/user_avatar/numba.discourse.group/esc/32/49_2.png) [@esc](https://numba.discourse.group/u/esc)\ **Post date:** [July 25, 2021, 8:14am UTC](https://numba.discourse.group/t/numba-0-54-0-rc2-and-llvmlite-v0-37-0-rc2/810/6 "2021-07-25T08:14:58Z")
Not explicitly no. But, you can of course use Dask to run Nuba accelerated functions, the Dask project even has an example page here: [Stencil Computations with Numba — Dask Examples documentation](https://examples.dask.org/applications/stencils-with-numba.html)