James Reinders

70 years ago this month, researchers ran the first FORTRAN program, considered the beginning of general-purpose programming: bit.ly/fortran7 Image v/@NASA

SYCL 2020 conformance is one of many new features in the Intel quarterly tools update. intel.com/content/www/us… @SYCLstd @IntelDevTools @IntelSoftware

New major rev on our software tools to help boost performance for AI, HPC, and more! intel.com/content/www/us…

Thanks to all who joined us @Supercomputing this year! It was especially great to see the enthusiastic turnout for the #SYCL Tutorial and BoF. There is an increasing number of developers using @SYCLstd, and we're excited to carry forward the momentum for heterogeneous computing

The IWOCL 2024 call for submissions is now open for international scientists and engineers from academia, industry, and government to contribute by presenting their work and discoveries using #OpenCL and #SYCL. iwocl.org @openclapi @argonne @thekhronosgroup

Intel Proposes Adding Full SYCL Programming Model Support To Upstream LLVM phoronix.com/news/Intel-Ful… via @phoronix

The 2nd edition of our SYCL book is now published and available (free pdf or epub). @SYCLstd link.springer.com/book/10.1007/9… eBook ISBN 978-1-4842-9691-2 Print ISBN 978-1-4842-9690-5

Unified Acceleration Foundation Forms to Drive Open Accelerated Compute and Cross-Platform Performance ⁦@SYCLstd⁩ linuxfoundation.org/press/announci…

Written by a professional SYCL Failure Taxonomy Intern - My Summer of Bad Code: an Investigation of Common Pitfalls When Writing SYCL Code ⁦@SYCLstd⁩ codeplay.com/portal/blogs/2…

Keeping the x86 ISA fresh and relevant for many decades is important for Intel and billions of users worldwide. APX is an important next step in our journey together IMHO. Looking forward to the journey together. intel.com/content/www/us… @intel @IntelSoftware #IamIntel

Intel tools update 2023.2, and exciting community news for open multivendor multiarchitecture accelerated computing. @intelnews @IntelSoftware @SYCLstd #IamIntel intel.com/content/www/us…

With Llama 2 release - we are sharing initial inference performance results for 7 billion and 13 billion parameter models on Intel's AI portfolio including Gaudi 2, Xeons, and PVCs. @intel @IntelAI #IamIntel intel.com/content/www/us…

SYCL is looking good. codeplay.com/portal/blogs/2… @SYCLstd #IamIntel

Intel oneAPI tools 2023.1 update all tools and libraries... further strengthen software development tools for open multivendor multiarchitecture heterogeneous/accelerated computing. intel.com/content/www/us… @SYCLstd @IntelDevTools @IntelSoftware #IamIntel

hpcwire.com/2023/02/13/an-… @SYCLstd @HPCwire #IamIntel

@illuhad @IntelSoftware @j_c_inPDX Great advice- the in order attribute for a queue that is in SYCL 2020 is an improvement we love over having different queue types! @SYCLstd #IamIntel

@IntelSoftware @j_c_inPDX How about the standard SYCL 2020 standard mechanism in_order queue for the same thing instead of unnecessarily using implementation-specific extensions like ordered_queue? 😉

New day, new question. Know the answer? Comment with the correct response and #oneAPISweepstakes for a chance to win a $20 USD gift card. Terms and conditions: intel.ly/3jSmntG #Sweepstakes

C++, #SYCL, #Kokkos – How C++ and Collaborative, Accelerator Programming Models Enable the Aurora Supercomputer @SYCLstd connectedsocialmedia.com/20234/c-sycl-k…

For those interested in how you can use #oneAPI and the new @codeplaysoft plugin to target multiple GPUs here's something for your end of year reading. Next year is getting more exciting as this starts to open up more possibilities! #programming @tonymongkolsmai/cuda-rocm-oneapi-running-code-on-a-gpu-any-gpu-28b7bf4cf1d0" target="_blank" rel="nofollow noopener">medium.com/@tonymongkolsm…

Another day, another opportunity for you to answer the question in the image. Just remember to include #oneAPI for a chance to win a $20 USD gift card! Terms and conditions: intel.ly/3v3GA2g #Sweepstakes

@codeandrew @tonymongkolsmai Orac wasn't really a computer.. it was more like a brain. Isn't that what they said or implied? I think that's why there's no LLVM backend for it.

When we say oneAPI works on multiple vendor architectures we mean it. Here’s the latest input to my test system. Intel Arc A770, AMD 7900XT and NVIDIA 3080Ti all living in harmony (thanks to a 1500W power supply). 😂😂😂