HP realizes that mandatory 15-minute support call wait times isn’t good support

HP rescinds European support call strategy due to "feedback."

arstechnica.com · Ars Technica

The Los Angeles Aqueduct is Wild — Practical Engineering

[Note that this article is a transcript of the video embedded above.] On the northern edge of Los Angeles, fresh water spills down two stark concrete chutes perched on the foothills of the San Gabriel Mountains, a place simply called The Cascades. It’s a deceptively simple-looking finish line: the

practical.engineering · Practical Engineering

Flash-KMeans: Fast and Memory-Efficient Exact K-Means

$k$-means has historically been positioned primarily as an offline processing primitive, typically used for dataset organization or embedding preprocessing rather than as a first-class component in online systems. In this work, we revisit this classical algorithm under the lens of modern AI system design and enable $k$-means as an online primitive. We point out that existing GPU implementations of $k$-means remain fundamentally bottlenecked by low-level system constraints rather than theoretical algorithmic complexity. Specifically, the assignment stage suffers from a severe IO bottleneck due to the massive explicit materialization of the $N \times K$ distance matrix in High Bandwidth Memory (HBM). Simultaneously, the centroid update stage is heavily penalized by hardware-level atomic write contention caused by irregular, scatter-style token aggregations. To bridge this performance gap, we propose flash-kmeans, an IO-aware and contention-free $k$-means implementation for modern GPU workloads. Flash-kmeans introduces two core kernel-level innovations: (1) FlashAssign, which fuses distance computation with an online argmin to completely bypass intermediate memory materialization; (2) sort-inverse update, which explicitly constructs an inverse mapping to transform high-contention atomic scatters into high-bandwidth, segment-level localized reductions. Furthermore, we integrate algorithm-system co-designs, including chunked-stream overlap and cache-aware compile heuristics, to ensure practical deployability. Extensive evaluations on NVIDIA H200 GPUs demonstrate that flash-kmeans achieves up to 17.9$\times$ end-to-end speedup over best baselines, while outperforming industry-standard libraries like cuML and FAISS by 33$\times$ and over 200$\times$, respectively.

arxiv.org · arXiv.org

젤렌스키 "미국서 21일 러우전쟁 종식 방안 논의 예정"

뉴시스 · [서울=뉴시스] 이재우 기자 = 미국과 우크라이나가 이르면 오는 21일(현지시간) 미국에서 러시아-우크라이나 전쟁 종전 방안 등을 의제로 회동에 나설 것으로 보인다.볼로디미르 젤렌스키 우크라이나 대통령은 19일 공개한 영상 연설에서 우크라이나 협상단이 러시아의 우크라이나 침략 전쟁 종식과 방공망을 포함한 안보 체계 구축 등을 논의하고자 미국으로 향했다고 밝..

www.newsis.com · 뉴시스

Warning to packagers: Minimum Go version bump to 1.26 is coming in two months · Issue #1525 · TecharoHQ/anubis

Hello, I want to increase the minimum Go version of Anubis from 1.24 to 1.26 so that among other things Anubis can benefit from the green tea garbage collector. If you package Anubis in a distro, p...

github.com · GitHub

The jbrains Experience: Affordable personal mentoring to help you start getting unstuck.

You need help, but you can't justify the expense of a full-time coach. Maybe you're not yet sure about buying one-on-one coaching sessions. You're struggling to convince your employer to pay for the mentoring you need. You need more than what the world offers you for free, but your budget doesn't have room for enterprise-level consulting. I would like to help you.

www.jbrains.ca · jbrains.ca

견주기: 의미를 견주어 단어를 찾는 게임

gyeon.postica.app

Behind the Blog: Marathon and the Metaverse

This week, we discuss unfortunately checking Twitter for news, the closure of the metaverse, and being vulnerable in Marathon.

www.404media.co · 404 Media

iOS Newsletter Issue 84

ios-newsletter.snappmobile.io