Profile avatar
axccl.bsky.social
computational physicist • accelerates accelerator research at @berkeleylab.bsky.social • laser-plasma aficionado • PByte connoisseur • mostly memory BW bound • openPMD • sciency C++ & 🐍 • GPGPU since before it was cool • he/his • Dr. rer. nat. Dipl.-Phys.
117 posts 263 followers 244 following
Regular Contributor
Active Commenter
Posts 20 Comments 30

All things, with one exception, have to end After years of sharing #HPC news and insights, it's time to log off It was fun while it lasted and I remain grateful to all of you who followed and interacted Stay curious, stay kind! — @hpcguru.bsky.social signing off #HPC #Farewell

submitted 19 days ago • 45 comments

MAJOR PARTICLE PHYSICS REPORT RELEASE ⚛️🧪 The National Academies particle physics decadal report is out! Read read! (I was a member of the committee that wrote this report but resigned during drafting bc of a conflict of service w/my HEPAP appointment.) nap.nationalacademies.org/catalog/2883...

submitted 19 days ago • 7 comments

HPSF 🤝 ISC 2025 Join us next week at ISC for tutorials, BoFs, and workshops on HPSF projects Spack, OpenCHAMI, and more. 🗓️ Details: hpsf.io/blog/2025/hp... #HPSF #ISC25 #ISChpc #ISC25_SCC #ConnectingTheDots

submitted 26 days ago • 0 comments

The inaugural High Performance Software Foundation Conference offered updates from HPSF projects, insights on processor & system trends, active discussions around the future of interoperability, and more. ▶️ Find our #HPSFCon recap and links to all the recordings at: hpsf.io/blog/2025/hp...

submitted 30 days ago • 0 comments

✨ Berkeley Lab's @nersc.bsky.social, @delltechnologies.bsky.social, and NVIDIA announce new public-private partnership for game-changing US Dept. of Energy Doudna supercomputer, (NERSC-10, strengthening U.S. capabilities, accelerating discovery, and powering future science. 🧪

submitted 31 days ago • 1 comment

NERSC-10 will be named Doudna: DOE awards contract to Dell for #supercomputer powered by NVIDIA’s next-generation Vera Rubin platform newscenter.lbl.gov/2025/05/29/d... #HPC #AI @nersc.bsky.social @cs.lbl.gov

submitted 31 days ago • 0 comments

new in pynucastro: we can now export to NetworkX and use their functions analyze a reaction network. Here's an example of finding cycles -- NetworkX finds the CNO and hot-CNO cycles in a network: pynucastro.github.io/pynucastro/n... #astro #astrocode

submitted 37 days ago • 2 comments

#HPC (as a developer) predominantly means trying to find strategies to get around limitations. You can't be root or completely control the environment. Your kernel or compiler is old. On a good day, it's still a few hours to get up the point of working to innovate on the thing you are interested in.

submitted 36 days ago • 1 comment

No API docs should have physics. Its super distracting.

submitted 37 days ago • 4 comments

Experimenting with something like this for @compiler-explorer.com - wondering what your thoughts are? AI is getting everywhere, and can be divisive. I'm hoping something like this - so long as I can get the output quality high enough - is a net benefit to help folks learn more about their code.

submitted 37 days ago • 21 comments

LLNL’s Center for Applied Scientific Computing (CASC) has developed a new JIT approach that leverages LLVM intermediate representation (IR) to optimize #GPU kernels at runtime computing.llnl.gov/about/newsro... #HPC #AI

submitted 39 days ago • 0 comments

📣 The 2026 INCITE Call for Proposals is open through June 16! The INCITE program supports groundbreaking research 🧪🔭🧬 requiring extreme-scale computing and awards allocations on Exascale systems including OLCF's #Frontier and ALCF's #Aurora 📝 More details: bit.ly/3VjoEhX #HPC #AI

submitted 39 days ago • 0 comments

Announcing the 44 Accepted #SC25 Workshops o Workshops with Proceedings o Symposium-style Workshops sc25.supercomputing.org/2025/05/anno... #HPC

submitted 40 days ago • 2 comments

I'm recruiting a PhD student to work on CMS, with a focus on substructure/BSM in four-top final states/CMS HGCAL: inspirehep.net/jobs/2921389 (all details in link)

submitted 44 days ago • 0 comments

Honored to be named a Vanguard of HPC + AI by insideHPC. At Argonne’s ALCF, I get to explore how humans and machines collaborate—through XR, in-situ viz, and tools that bring data to life. Dancing on the bleeding edge of science—and loving every step. insidehpc.com/2025/05/vang...

submitted 47 days ago • 0 comments

HPSF Conference Day 3️⃣ starts off with a bang 🧨 Today's #HPSFCon agenda features in-depth sessions for several HPSF project communities including: 🎇 AMReX 🎇 Apptainer 🎇 Charliecloud 🎇 Kokkos 🎇 Spack 🎇 Trilinos

submitted 54 days ago • 1 comment

We're looking forward to HPSFCon next week with more than 130 speakers on the most pressing high performance software topics 🔥 Featured panels include: 👥 Status and Trends in the HPC Landscape 👥 Processor Trends and What They Mean for Software Join us in Chicago to be part of the conversation!

submitted 59 days ago • 2 comments

Pro tip: even on #conda-forge, CI/CD ain't no beauty contest. (RE: disabling tests/entries w/o opening an issue to track.)

submitted 68 days ago • 0 comments

Discussion of *why* AI hardware as a field prefers low precision types came up (again) the other day. The software brained answer is that for precision insensitive workloads, you effectively double your bandwidth from caches/device mem/host mem But there's a lot more to it.

submitted 72 days ago • 1 comment

Are you also frantically working to make your #SC25 paper as rigorous as you can? Check our scientific benchmarking guidelines: buff.ly/s0PjSoW 🎯. It's their 10th anniversary since #SC15 and still going strong 🥂 . Happy writing to all #HPC and #AI folks 🚀! #reproducibility

submitted 80 days ago • 0 comments