With Infinia gaining support for files, is there a reason for anyone to buy EXAScaler anymore? Why would someone choose Lustre over Infinia if Infinia seems to support everything that Lustre does but with enterprisey glitter? Curious to see how DDN positions these two lines. #HPC

Block off your calendar for Wednesday at 10am! I'll be there.

=> Architecting for Frontier AI Labs: How Anthropic scales Claude on TPUs, Google & Anthropic, Google Cloud Next 2026, Apr 22 youtube.com/watch?v=LEPY... content-cdn.sessionboard.com/content/3e8V... Automated fault recovery & intelli scheduling to maintain continuous uptime across massive pod-scale

I wrote up some notes after my week at #ISC26: blog.glennklockwood.com/2026/07/isc2... Interested to hear others' perspectives on highlights and things I may have missed. #HPC

I somehow missed that China has developed its own implementation of InfiniBand, called ScaleFabric 400, to compete/supplant NVIDIA/Mellanox IB. Oddly enough, its manufacturer (Sugon) is not a member of IBTA though. en.eeworld.com.cn/mp/EEWorld/a... #HPC

Interesting that C-DAC, whose indigenous Indian interconnect Trinetra is based around libfabric, opted to implement verbs support to make Lustre work instead of adopting Cray’s kfabric LND. #ISC26

1. General LLMs beat coding-specific models at writing Fortran, likely bc code models are overfitted to Python 2. This conf begins the age of Claude-generated pptx. Clear tells 3. My phone doesn’t like these #ISC26 projectors

Now happening at #ISC26 #HPC

This is the worst Jensen keynote ever

I can’t wait to hear what Jack Dongarra thinks about the future of #HPC at #ISC26’s closing keynote.

Dumb stuff that results in weird requirements in #HPC RFPs. Ultra dense GPU racks with all liquid requires unusual datacenter construction considerations. #ISC26

Ogawa-san is how I first came to know about China’s new exascale system a while ago, and I assembled a quick spec page a few weeks back (glennklockwood.com/garden/syste...). Notably has a BIG liquid cooled storage system (67 racks, 10 TB/s).

??? I was the lowest-level principal software engineer at microsoft and my parting annual liquid total comp started with a 4 when I quit. It was enough to support two of us on a single income in the Bay Area. Priced out of Redmond??? Sorry, but not very sympathetic.

It’s comforting to know that, at the macro scale, Lustre is pretty consistent at its strengths and weaknesses. I don’t think these are different from when I last used Lustre in 2022. From @adrianjhpc.bsky.social at the Lustre BOF in #ISC26

lol James Lin at #ISC26 #HPC

Is this because other energy companies don’t have multiple big supercomputers? Or because they don’t list them if they do? Fact that Exxon doesn’t go in as hard as ENI suggests either energy doesn’t need them

Hot #ISC26 tip: VAST is running free taxi service from the taxi stand in front of CCH to the SIDE hotel downtown and back. If you don’t feel like getting sweaty before/after an offsite meeting, I think anyone can hop in. Look for the city cabs with the VAST logo on the side.

The jump of LineShine over El Cap in efficiency is staggering. Suggests the benefit of balance and design for FP64-first over wedging FP64 into a socket whose chiplets were AI-first. LineShine’s poor HPL-MxP underscores. #ISC26 #HPC

“Heterogeneity is here to stay.” Also, “the new #1 system is homogeneous all-CPU.” #ISC26 #HPC

China’s back at the top of Top500. I guess the threat of not being allowed to buy US-designed chips is no longer a compelling reason not to submit top systems. Happy 250th, USA. #HPC #ISC26

20 minutes into the opening keynote at #ISC26 and Martin Schulz has managed to say AI only once; it’s changing everything, but let’s talk about quantum and #HPC instead. Impressive.

The #HPC (+ #AI ) Leadership Training Institute returns to Barcelona: 2026 Sept 7-10. Strategy, technology evaluation, cost models, TCO, procurement, acceptance testing, sovereign HPC/AI, balancing on-prem / cloud / AI, business cases, + more. Registration & more info soon!

I think if you replace “Everpure’s AI strategy” with “the global economy,” the headline sounds about right www.nextplatform.com/store/2026/0...

what the heck is an SEO blog?

if there wasn’t a bubble before…

The sense of entitlement from the highly paid employees at these big tech companies is flooring. If these companies treat you so badly, why don’t you just work at a better company? It’s what I did. There’s no free lunch in life. www.wired.com/story/mark-z...

How mad is Sam right now that his model’s never been export controlled?

I want a pony

🎓 Last day to register and Help Shape the Next Generation at #ACM #PEARC26! You don't need the same research background: help students navigate the conference, build networks, and explore career paths! Interested? Email [email protected] or update your registration. #RCD #PEARC #HPC #student

This interesting—a hardware dot product ALU in the HBM stack to accelerate KV cache. Puts computation behind the skinny wire that connects memory to ALUs in conventional accelerators.

"DDN is seeking to attract investors that can help the company better reflect the value it creates" One of the weirder statements I've seen recently. Implies that current investors do not value their investment as highly as the investment values itself. www.bloomberg.com/news/article...

Was reminded on @hpc.social that PEARC26 regular registration closes this Friday! It's a great, practitioner-focused #HPC and #RCD conference, and I will be there this year. pearc.acm.org/pearc26/regi...

"FP8 is all you need" vs "A naive Ozaki II implementation with r=10 residue planes per warp can spill, raising β above 1 and erasing the projected speedup." Which one wins? #HPC

The latest issue of the Chapel newsletter is out now, featuring updates on Chapel funding, talks, papers, interviews, the website, and more! Check it out at chapel.discourse.group/t/chapel-new...

I am convinced there is a growing lexicon of AI terms that are used exclusively by people who don't understand AI to talk to other people who don't understand AI. For example, has any real AI practitioner ever referred to "keeping GPUs fed?" #AI

Microsoft AI's report on MAI-1 training is full of interesting infrastructure anecdotes, esp around storage: - Async checkpoint from DRAM to object (no node-local SSD) - Restart is read once, then broadcast - Anyscale Ray for fast restart - Bitwise repro > MFU microsoft.ai/wp-content/u... #AI

Virtual DAOS User Group slides just dropped (daos.io/event/virtua...). Interesting content, and glad DAOS is still trucking along. Surprises: ALCF has only 37% servers in production; LinkedIn trying to run DAOS on 20% of cores on HGX servers; long feature roadmap. #HPC #storage

📢 [CFP] PDSW 2026 Workshop at SC26! I'm happy to announce the 11th International Parallel Data Systems Workshop, to be held on Nov. 16, 2026 in conjunction with SC26! Submissions are due July 31, 2026 (11:59PM AoE). 🤩📜 Additional information: www.pdsw.org #HPC #supercomputing

Fugaku-NEXT design report dropped. Amazing details of a bold vision in here. Job scheduling by both k8/s (for #AI) and Slurm/PBS/Flux (for #HPC), >100 TB/s two-tier file system. VAST, WEKA, and Scality(!) named as interesting technologies for all-flash tier. www.r-ccs.riken.jp/outreach/top...

I got to work with VAST's architects on integrating NVIDIA Vera BlueField-4 STX in-silicon security features with VAST's agent policy enforcement engine. Agent interactions can be inspected and governed at line rate, and policies can adapt in real time. Very slick. www.vastdata.com/blog/vast-ze...

Claude offered to create an SVG of Europe for me (I did not ask), and this is what came out. We have a long ways to go before the Terminators take over.

I am living in the future, and the future is being a human who finds all the problems in AI-generated technical slop. Even as models/agents get better, the humans using them aren't. So they still produce nonsense, albeit shinier nonsense, and I remain employed. #AI

Found my new LinkedIn profile while visiting my mom this weekend

I think there is a typo in the headline. “up to 90:1 compression card.” That aside, I’m glad to see others adopt more sophisticated transformations in their data reduction algorithms to distill out more entropy.

Today's revelation: there is a ton of overlap between writing aspirational grant proposals and writing marketing copy for aspirational technology partnerships. Academics who look down their noses at marketing and sales are lying to themselves about what skills they are using to get funding.

New PNAS paper finds that GPT-4.5 was more human than humans when prompted correctly, passing the Turing test. Alternate headline: Turing test is proven to be not very useful. Pop news: www.the-independent.com/tech/ai-turi... PNAS paper: www.pnas.org/doi/10.1073/... #AI

A career in collaborative grant writing, writing papers, and peer review has prepared me for spending days (weeks?) of scrubbing deep technical marketing copy.

I just got confirmation that my login credentials on NERSC supercomputers were revoked at some point and no longer work. With my Microsoft free Azure gravy train also ending late last year, I am officially without a supercomputer. First time in my career 😢

I don't know about the other parts, but it appears that the storage for this only scales up to 16 PB (see www.delltechnologies.com/asset/en-us/...). I'm surprised that's being positioned as scalable AI infrastructure. #storage

The virtual DAOS User Group meeting is this Thursday (May 21) starting at 7am Pacific. Check out the speaker lineup and agenda here: daos.io/event/virtua... #storage #AI #notsosureaboutHPC