AI and performance consulting for technical teams. CUDA, OpenCL, oneAPI/SYCL, C/C++, Python, edge systems, and multi-GPU workloads.
Atlanta, GA
Joined December 2008
- Tweets3.1K
- Following3K
- Followers9K
- Likes240
ArrayFire retweeted
Reading through the @arrayfire source code to learn how to juggle between multiple compute backends and I'm having a blast. Code is well-formatted and well-documented and it's quite easy to just 'follow along'. Bravo @arrayfire team!
ArrayFire retweeted
Getting an update on #G2GPU from Natan Biesmans at the Dutch LabVIEW User Group meeting today. A library in the making to make using ArrayFire in #LabVIEW a breeze!
#dutlug #labviewfriends
ArrayFire retweeted
Congratulations @melonakos and the @arrayfire team on joining @intel’s #OneAPI effort
intel.com/content/www/us/en/…
#HPC #AI #GPU
Today, we are announcing that my team and I have joined @sanjivmshah and team @Intel to open a new chapter in our story. We look forward to contributing to oneAPI and its open-source vision for accelerated computing.
Check out the blog post below for more!
ArrayFire retweeted
Intel Arc Pro GPUs is a next step in providing open, end-to-end solutions to scale performance like the Intel Open VKL plugin for @pixarsrenderman, the Intel Open Path Guiding Library (OpenPGL), and more. Learn more: intel.ly/3djmRWK
#SIGGRAPH2022 #oneAPI @Blender
ArrayFire retweeted
The Parallel Universe, Issue 48 is out! Featuring the NeurIPS Billion-Scale Approximate Nearest Neighbor Search Challenge, additions to #SYCL 2020, #AI pipeline optimizations, and more! intel.ly/3zPhOXj
#oneAPI @IntelAI @arrayfire @SYCLstd
ArrayFire retweeted
.@arshad_umar from @ArrayFire and @HenryGabb2 from @Intel, compare two approaches to accelerate the 2D Fourier correlation algorithm for heterogeneous parallelism with #ArrayFire and #oneAPI. Take a look here: intel.ly/3GdnM5r
ArrayFire retweeted
Read the latest from @codeproject about the #oneAPI Deep Neural Network (oneDNN) library using #SYCL based programming.
@SYCLstd @arrayfire @openclapi
ArrayFire retweeted
Check out 3rd-party wrappers enabling transparent NumPy functions with @arrayfire. #PyConUS2022 #OpenTeamsPartners. pypi.org/project/cocos/ | pypi.org/project/afnumpy/
ArrayFire retweeted
Our partner, @arrayfire offers a library to accelerate your code for parallel computing with an easy-to-use API. Check out #Python wrappers bit.ly/opensource-wrappers | arrayfire.org/arrayfire-pyth… (documentation) #PyConUS2022 #OpenTeams
ArrayFire retweeted
OpenTeams’ partner, @arrayfire has built the fastest library for GPU computing. Get started with their library in #Python in under a minute on Google Colab bit.ly/arrayfire. Want to know more? Visit us at booth 440, exhibition hall, #PyConUS2022
ArrayFire retweeted
#oneAPI greatly simplifies development on heterogeneous accelerators. Integrate oneDNN Library and #SYCL into @arrayfire code bases, when migrating to a new programming model. Learn how: intel.ly/3iFn2uW
#deeplearning #Developers @thekhronosgroup @SYCLstd
ArrayFire retweeted
Explore how to integrate the #oneAPI Deep Neural Network (#oneDNN) library for higher levels of heterogeneous parallel abstraction in @arrayfire. intel.ly/3LRC0LR
@SYCLstd
ArrayFire retweeted
Conway's Game of Life on GPU using 2D convolutions, a sweet experiment in Rust using ArrayFire. Hold your hats gentlefolks as minds will be blown.
link.medium.com/9LFh3cJRlnb #Rust #ArrayFire #Medium
ArrayFire retweeted
Simple machine learning inference is usually anything but. Thankfully, #oneAPI plays its part to ensure that simplicity. Join Stefan Yurkevitch of @ArrayFire for this on-demand #DevSummit session to learn more: intel.ly/3ds3tDK
ArrayFire retweeted
Plenty claim "#exascale" based on only single-precision peak theoretical performance or by "combining" compute power to get there.... "neither of these count, at least by the standard definition." nextplatform.com/2021/09/23/…
@TheNextPlatform @NicoleHemsoth #HPC #supercomputing
ArrayFire retweeted
Our recent article “Cross-Platform GPU-Based Implementation of Lattice Boltzmann Method Solver Using ArrayFire Library” has been published in @MathematicsMDPI! @arrayfire #LBM #LatticeBoltzmann #CFD #GPGPU #CUDA #OpenCL #D2Q9 #D3Q27
doi.org/10.3390/math9151793 – at Slovak Republic
ArrayFire retweeted
Our paper on the 1. GPU implementation of the CCM algorithm is on @IEEEXplore Run on ABCI where we speed up CCM 1530X by converting memory functions into arithmetic functions with @arrayfire to make it GPU compatible. #DataScience #BigData
ieeexplore.ieee.org/document…
ArrayFire retweeted
We can't wait for this live #DevToolsChat with the #OpenMP and #SYCL communities, and we're excited about the special guests joining us. Don't miss it! Next Wed, June 2, 9 am PT.
@arrayfire @BristolUni @codeplaysoft @OpenMP_ARB @SYCLstd @URZINFO @IntelSoftware
ArrayFire retweeted
Currently implementing the interop between @arrayfire (CUDA+OpenCL) and DirectX (since @unity3d’s XR Plugin doesn’t support OpenGL textures for Oculus Rift).
Looking for good tutorials on CUDA/OpenCl /w DirectX interop - can somebody recommend any?
Interactive Lattice-Boltzmann simulation visualized in #VR (Oculus Rift), updating in real-time. Only 2D output is implemented (this way the immersiveness of VR cannot be fully utilized). Already written the 3D sim code, but not yet implemented in VR prototype.