Lightnews — Scholar-powered news

Rob Mueller-Albrecht @roboima.bsky.social · Jul 14

Open multiarchitecture acceleration frameworks powering #AI everywhere, underpinning #PyTorch for #ARM, #Intel, #x86, #NPU, and #GPU with portable APIs.

Check out this presentation from the #UXL Foundation Mini-Summit

youtu.be/RxYIhUctlvA?...

How the UXL Foundation Projects are Powering AI – Megan Knight & Rod Burns

YouTube video by UXL Foundation

youtu.be

1 1

Rob Mueller-Albrecht @roboima.bsky.social · Jul 9

The latest Intel CPU optimizations on an open-source foundation with LLVM backend: Leverage the Intel® oneAPI DPC++/C++ Compiler to build your custom Linux kernel.
#oneAPI #Linux #CPP #LLVM

Learn how to set it up and what to consider:
www.intel.com/content/www/...

Build Linux* Kernel with Intel® Compilers

Use the latest Intel® oneAPI DPC++/C++ Compiler to build a Linux kernel. Take advantage of the latest Intel hardware features, compiler optimizations, and LLVM Clang capabilities.

www.intel.com

Rob Mueller-Albrecht @roboima.bsky.social · Jun 30

Scalable multiarchitecture parallelism unleashed with the 2025.2 tools release!

Scale and optimize any workload. One solution for AI, HPC, and imaging through seamless interoperability between #MPI, #OpenMP, #SYCL, #PyTorch, #Vulkan, #DirectX, and ...
#oneAPI #AIPC

www.intel.com/content/www/...

Developer Tools 2025.2 Parallelism Unleashed

Scale, optimize, and streamline the execution of any workload through comprehensive, highly integrated parallelism, on an AI PC or a supercomputer. One developer solution for AI, HPC, and imaging.

www.intel.com

2

Rob Mueller-Albrecht @roboima.bsky.social · Jun 30

Supercharge parallelism, with optimized performance and productivity for AI, graphics, and accelerated compute.

Intel #oneAPI Toolkits & AI Tools 2025.2 are here, delivering faster AI inference, real-time rendering, and expanded HPC support.

#Xeon #Core #IntelAI

www.intel.com/content/www/...

Intel® Software News Updates

Stay up to date on the latest software happenings, including software releases, new capabilities, ecosystem adoption, success stories, and more.

www.intel.com

Rob Mueller-Albrecht @roboima.bsky.social · Jun 23

Imagine AI seeing and hearing how you feel and responding with empathy and understanding?

#LAION in collaboration with #IntelAI just released EmoNet — a suite of ready-to-use #OpenSource tools, models, and benchmarks.
#emotionAI #HuggingFace

www.intel.com/content/www/...

LAION - Do They See What We See?

Building Emotionally Intelligent AI with EmoNet, a suite of open tools and resources developed by the LAION community.

www.intel.com

3

Rob Mueller-Albrecht @roboima.bsky.social · Jun 23

Something my husband's daughter (neuro-oncologist & palliative care specialist) shared with me. Please forward if you know someone who may get useful insights from this podcast:
geripal.org/aging-and-tr...

Transgender Health, Aging, and Advocacy: A Podcast with Noelle Marie Javier and Jace Flatt

On today's episode we discuss Aging and Transgender, including issues older transgender individuals face, with guests Noelle Marie Javier and Jace Flatt.

geripal.org

Rob Mueller-Albrecht @roboima.bsky.social · Jun 16

Find out about the latest Intel contributions to #GCC 15. Get optimized performance out of the #Intel #Xeon 6 processor with E-Cores and P-Cores, and future Intel Xeon and Intel Core Ultra processors.

Take full advantage of Intel #APX with our open source commitment.

www.intel.com/content/www/...

Next Generation Optimizations for GCC* 15

Bringing the Next Generation of Intel® CPU Performance Optimizations to GCC* 15. Full Implementation of Intel® Advanced Performance Extensions

www.intel.com

1

Rob Mueller-Albrecht @roboima.bsky.social · Jun 3

#IXPUG: Scale parallel offload acceleration from laptop to exascale supercomputing with the latest #OpenMP features and enhanced #OpenMP composability to ensure seamless highly performant multi-node parallelism between #MPI, OpenMP and #SYCL.

www.intel.com/content/www/...

OpenMP* and the Future of Scientific Computing

OpenMP* and the Future of Scientific Computing.. Composability facilitates the use of SYCL libraries together with OpenMP and MPI with minimal changes.

www.intel.com

1

Rob Mueller-Albrecht @roboima.bsky.social · Jun 3

#IXPUG : Leverage hierarchical parallelism and strategic workload allocation on CPU & GPU with #OpenMP and #MPI. Achieve deep understanding of light-matter interaction faster by applying #oneMKL mixed precision with BF16 for #HPC and not just #AI.

www.intel.com/content/www/...

Deepen the Understanding of Light-Matter Interaction

Deepen the Understanding of Light-Matter Interaction. Accelerate high-performance computing HPC with mixed precision oneMKL, OpenMP and MPI

www.intel.com

1

Rob Mueller-Albrecht @roboima.bsky.social · Jun 3

#IXPUG Unified Joint Matrix #SYCL Extensions in #oneAPI tools provide a #SPIRV compatible path to high performance open multiarchitecture abstraction for diverse matrix hardware.

Get the best #AI & #HPC offload performance
#XMX #AMX

www.intel.com/content/www/...

Unified Joint Matrix SYCL* Extension

What if we had a unified interface for matrix and tensor operations compatible with SPIR-V cooperative matrix extensions, made it part of SYCL, and thus made it available to a wide variety of hardwar

www.intel.com

1

Reposted by Rob Mueller-Albrecht

Dubois County Democratic Party @duboiscountydems.bsky.social · May 23

Tuesday, May 27 at 6:00 PM
Agaves Mexican Grill, 1935 Main St, Ferdinand, IN
Guest Speaker: Ken Sicard, Ferdinand Town Council

How do TIF (Tax Increment Financing) districts work?
How do they impact local budget & taxes?
Engage with fellow community members in a relaxed setting.
See you there!

1 1

Rob Mueller-Albrecht @roboima.bsky.social · May 20

Accelerate bindless image processing and multi-architecture offload kernel memory access with #SYCL #Vulkan and #DirectX 12 interoperability, using the latest Intel DPC++ Compatibility Tool or the Intel #oneAPI DPC++/C++ Compiler.

www.intel.com/content/www/...

SYCL Interoperability with DirectX and Vulkan

Vulkan and DirectX12 interoperability fir bindless image support minimizes the amount of data that has to be transferred back and forth between CPU host processes and accelerator devices.

www.intel.com

Rob Mueller-Albrecht @roboima.bsky.social · May 20

#SYCL continues to evolve. Discuss, learn, and experience the use of C++ virtual functions with C++ and SYCL for offload kernels. Find out how the latest implementation updates of the #SPIR-V backend for #LLVM compilation with #SYCL code benefits you.

www.intel.com/content/www/...

Dimensions of SYCL* Expansion: C++ Virtual Functions and More

Dimensions of SYCL Expansion: C++ Virtual Functions and SPIR-V Backend

www.intel.com

Rob Mueller-Albrecht @roboima.bsky.social · May 11

www.hpcwire.com/off-the-wire...

Argonne Leverages AI and Supercomputing to Advance Cancer Research

LEMONT, Ill., April 22, 2025 — Discovering new drugs to treat cancer and predicting how tumors will respond to them remain key challenges in the fight against the disease. At the […]

www.hpcwire.com

Rob Mueller-Albrecht @roboima.bsky.social · May 11

www.hpcwire.com/2025/04/24/n...

Nature Reports a US Science Brain Drain Has Begun

A report in the prestigious science journal Nature suggests a brain drain has started in the U.S. with fewer scientists applying for positions in the U.S. and an increase in […]

www.hpcwire.com

Rob Mueller-Albrecht @roboima.bsky.social · May 2

Dive deeper into Intel Open Image Denoise's Scientific and Technical Award from the Academy of Motion Picture Arts and Sciences.
Learn how its open standards-based approach benefits animation and visual effects.

#CGI #oneAPI #CPP #SYCL #PyTorch

www.intel.com/content/www/...

Open Image Denoise Scientific and Technical Award

Dive into the Intel® Open Image Denoise Scientific and Technical Award for Movie Rendering Acceleration.

www.intel.com

1

Rob Mueller-Albrecht @roboima.bsky.social · May 2

Intel Open Image Denoise Wins Scientific and Technical Achievement Award from the Academy of Motion Picture Arts and Sciences. Find out more about its contributions to the movie industry:

newsroom.intel.com/intel-produc...

#oscar #rendering #raytracing #denoising

Intel Open Image Denoise Wins Scientific and Technical Achievement Award

newsroom.intel.com

1

Rob Mueller-Albrecht @roboima.bsky.social · Apr 16

#SYCL is growing ever more powerful, supporting multi-architecture accelerated compute for #Edge, #HPC, and #AI use cases on more and more diverse devices.

Check out the State of the Union keynote from #IWOCL 2025.

youtu.be/qZSbqbFNmpU

SYCL State of the Union

YouTube video by IWOCL

youtu.be

2

Rob Mueller-Albrecht @roboima.bsky.social · Apr 9

Intel's Parallel Universe Magazine Issue 58 is now available! 🎉

Discover how Intel® oneAPI, Intel® Extension for Scikit-learn, Intel® Tiber™ AI Cloud, and Intel® GPUs are driving innovation.

📖 Read the full issue: intel.ly/42jnyFg

Rob Mueller-Albrecht @roboima.bsky.social · Apr 7

Anticipate financial risks and model option and derivative valuation in challenging times.

Rob Mueller-Albrecht @roboima.bsky.social · Apr 7

Model, understand and predict financial market trends leveraging #oneAPI, #SYCL, #oneMKL for accelerated math and numerical differential equation solution.
#stocks #finance #greeks

A quick overview of how it is done:
www.intel.com/content/www/...

Accelerate Quantitative Finance with SYCL and oneMKL

Explore prediction models for market valuation and how they benefit from running an optimized random number generator on parallel compute engines.

www.intel.com

Rob Mueller-Albrecht @roboima.bsky.social · Apr 7

Model, understand and predict financial market trends leveraging #oneAPI, #SYCL, #oneMKL for accelerated math and numerical differential equation solution.
#stocks #finance #greeks

A quick overview of how it is done:
www.intel.com/content/www/...

Accelerate Quantitative Finance with SYCL and oneMKL

Explore prediction models for market valuation and how they benefit from running an optimized random number generator on parallel compute engines.

www.intel.com

1 3

Rob Mueller-Albrecht @roboima.bsky.social · Apr 7

Keynote “The Evolution of Developer Software” by Sanjiv Shah, Intel, and talks by many other experts in tuning and optimizing AI and HPC workloads and applications. Join us online April 15! ixpug.org/events/ixpug... #CPU #GPU #oneAPI

IXPUG Annual Conference 2025 | IXPUG

Join the IXPUG Annual Conference 2025, hosted by TACC, for expert insights. Connect with leaders and explore innovations.

ixpug.org

1

Reposted by Rob Mueller-Albrecht

MADVoters @madvoters.bsky.social · Apr 3

The impact of 20 years of dwindling investment in education and infrastructure in our state has collided with the current presidential administration's scorched earth policies - and Hoosiers are the ones who will suffer.
www.wishtv.com/news/indiana...

Stellantis announces layoffs at Kokomo plant

Stellantis temporarily lays off 900 workers due to President Trump's new auto tariffs. Production halted at Canadian and Mexican plants.

www.wishtv.com

1 5 6

Rob Mueller-Albrecht @roboima.bsky.social · Apr 4

Very insightful:
Responsible AI use remains a top priority across the public sector to ensure ethical, lawful, and mission-appropriate deployment—especially in high-stakes environments like the DOD, where AI decisions can have life-or-death consequences.

battle-updates.com/meeting-the-...

Meeting the High Threshold in DOD for Responsible AI Use By Rear Adm. Ret Ron Fritzemeier, Director of Mission Solutions, Intel

battle-updates.com

1

Rob Mueller-Albrecht @roboima.bsky.social · Apr 3

Have you ever been stuck doing
$ make clean; make
app rebuilds, that take too long?

The Intel #oneAPI DPC++/C++ Compiler 2025.1 now comes with #ccache support, accelerating rebuilds and increasing productivity. It also works for #SYCL code!

Find out more:
community.intel.com/t5/Blogs/Tec...

Speed-Up C++ and SYCL* Application Recompilation Cycles with Ccache*

One pain point in maintaining a large code base is that even the smallest code change can lead to long recompilation times when significant dependencies exist throughout an application. If only there ...

community.intel.com