Listening to #SCAsia / #HPCAsia closing session keynote by Kathy Yelick on how the community of #HPC must adapt in the new world of computing driven by #AI …
#HPC
I had a wonderful time at #HiPEAC26 in Kraków! It was great to connect with the European #HPC and cloud community. We presented our work on Cppless - #serverless functions as first-class citizensin C++ 🚀
Paper ➡️ https://spcl.inf.ethz.ch/Publications/index.php?pub=508
Code ➡️ https://github.com/spcl/cppless
More tips on #HPC for improving your utilisation whether you’re running scientific compute, quants analytics or AI or pretty much anything else.
cloudhpc.news/improving-ut...
Improving Utilisation in HPC a...
Amirreza Rastegari - benchmarking expert from Microsoft #HPC / #AI engineering group - giving talk at #SCAsia / #HPCAsia on Azure HBv5 clusters using AMD CPUs with HBM plus InfiniBand interconnect - enabling sector-leading #HPC application performance via huge memory bandwidth
Bàn về hiệu năng hệ thống AI workstation kép RTX PRO 6000 với 1.15TB RAM: So sánh xử lý GPU-only (INT4) vs CPU+GPU (fp8) trên mô hình MiniMax-M2.1. Kết quả: GPU-only nhanh hơn 2–4x ở prefill nhưng chỉ xử lý tối đa ~3 request đồng thời do giới hạn KV-cache..fp8 tuy chậm hơn nhưng mở rộng tốt hơn cho 10+ người dùng, đặc biệt với context dài. Queue time là điểm nghẽn quan trọng. Phù hợp cho agent coding nội bộ. #AIWorkstation #LLMBenchmark #MultiUserAI #GPUvsCPU #LocalLLM #HPC #MachineLearning #Tín
Just a few weeks ago I read a blog post by someone recapping how multiple package management systems have tried and failed to use git as their backends, and just today I am trying to set up spack (https://github.com/spack/spack) for our #HPC cluster, and I'm not sure how I feel about it yet. I feel slightly burned by systems that you set up by git-cloning them.
I mean, apart from the quite, uh, unfortunate connotation of the name in a German-speaking context.
Wussten Sie, dass auch die Entwicklung von Wissenschaftscodes fürs #HPC gefördert werden? In Bayern kümmert sich #KONWIHR darum und bis zu 50.000 Euro sowie Expertise vom LRZ oder vom RRZE in Erlangen. und hat in KONWIHR steht für #Codes, #Libraries und #Tools wie waLBerla, preCICE, MARQOV oder dealii. Wie KONWIHR funktioniert, erklären Gerhard Wellein von der @FAU sowie Hans-Joachim Bungartz von der @tu_muenchen https://www.lrz.de/news/detail/konwihr-kein-geld-verplempern.
For graduate students and postdoctoral scholars working in #HPC, computational science and data-driven research. Application for the 16th International High Performance Computing Summer School is open 𝘂𝗻𝘁𝗶𝗹 𝟯𝟭 𝗝𝗮𝗻𝘂𝗮𝗿𝘆: ss26.ihpcss.org #HPCCS26 will take place on 12-17 July in Perth, Australia 🇦🇺
New #HPC Job! 💻️: Title: HPC Systems Administrator
Location: Houston, TX
Employer: Rice University
Remote: Hybrid
https://emdz.fa.us2.oraclecloud.com/hcmUI/CandidateExperience/en/sites/CX_2001/job/5464
New #HPC Job! 🤩️: Title: Computational Scientist for AI
Location: Houston, TX
Employer: Rice University
Remote: Hybrid
https://emdz.fa.us2.oraclecloud.com/hcmUI/CandidateExperience/en/job/5465
New #HPC Job! ⭐️: Title: Computational Scientist for AI
Location: Houston, TX
Employer: Rice University
Remote: Hybrid
https://emdz.fa.us2.oraclecloud.com/hcmUI/CandidateExperience/en/job/5465
One of the things that is making me look forward to 2026: the UK Government has announced this morning an upgrade for our supercomputing facilities at the University of Cambridge @Cambridge_RCS
Having done this a few times now I know what we in the team are in for - getting it installed, benchmarked, and in to service. But what a great asset for the Cambridge research community and via #AIRR the whole of the UK!
https://www.cam.ac.uk/news/government-funding-boost-for-cambridge-supercomputer
@BenjaminHCCarr
HPC & AI Review 2025: "Rise of AI for Science"... but academia still fears speculative generative models? 🤔
**Rigorous work** (TMST threshold, Teorema 4.3.1): Analytic entanglement dominance T→0
**Speculative work** (MVC vortex-QCD): Confinement via collective TMSS projection
Too "speculative" for some instances (*cough*), yet both codes run on any laptop. 🤷♂️
**Code > Credentials**
https://github.com/JavierMartinAlonso1980/entanglement-dominance-tmst
Get ready to understand how to use Linaro Forge to Optimize performance, when migrating workloads between hardware architectures.
Our experts will be leading the talk on 📍 29th January @13:20 JST
📍Booth S-9
📅Until Jan 29th
#2025: An #HPC Year In Review
Trump #AI Regs, #DOGE Cuts, and Gov’t Shutdown
#DeepSeek Ushers In Reasoning Models and #AgenticAI
Rise of AI for #Science
#TrillionParameterConsortium (#TPC25) Clears the Tower
The #China Drama Continues
(A)I scratch Your Back, You Scratch Mine
New #Supercomputers
New Chips, Faster Bits
#Quantum Leaps Forward
Power and Cooling
https://www.hpcwire.com/2025/12/17/2025-an-hpc-year-in-review/
Tạo động cơ LLM 1.58-bit chạy 117 token/giây trên 1 nhân CPU với Rust và AVX-512, nhưng bị lỗi ở lớp Activation khiến đầu ra luôn là <unk>. Cần hỗ trợ về: (1) Weight tying trong BitNet – thiếu hệ số tỉ lệ? (2) Cách scale tích lũy nguyên từ VPOPCNTDQ trước khi đưa vào RMSNorm/SiLU. Dự án mã nguồn mở, zero-copy, không heap allocation. #Rust #AVX512 #LLM #MachineLearning #AI #R3Engine #BitNet #LocalAI #HPC #Inference #trítuệnhân tạo #môhìnhtonngẫu #xửlýsongsong #tinhoccao
https://www.reddit.
Only a week to go till the #FOSDEM2026 #HPC, #BigData, and #DataScience devroom in Brussels! Check out the agenda and options for information on how to connect at fosdem.org/2026/schedul...
FOSDEM 2026 - HPC, Big Data & ...
Client Info
Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst