Arch Linux: Recent news updates

The following packages may require manual intervention due to the upgrade from 9.0 to 10.0:

● aspnet-runtime
● aspnet-targeting-pack
● dotnet-runtime
● dotnet-sdk
● dotnet-source-built-artifacts
● dotnet-targeting-pack

pacman may display the following error failed to prepare transaction (could not satisfy dependencies) for the affected packages.

If you are affected by this and require the 9.0 packages, the following commands will update e.g. aspnet-runtime to aspnet-runtime-9.0:

pacman -Syu aspnet-runtime-9.0

pacman -Rs aspnet-runtime

source
(author: George Rawlinson)

Chips and Cheese
SC25: Estimating AMD’s Upcoming MI430X’s FP64 and the Discovery Supercomputer
#ChipAndCheese

Telegraph | source
(author: George Cozma)

Telegraph

SC25: Estimating AMD’s Upcoming MI430X’s FP64 and the Discov…

Hello, you fine Internet folks, At Supercomputing 2025, EuroHPC, AMD, and Eviden announced that the second Exascale system in Europe was going to be called Alice Recoque. Powered by Eviden’s BullSequena XH3500 platform using AMD’s upcoming Instinct MI430X…

Daniel Lemire's blog
Why speed matters

Telegraph | source

Telegraph

Why speed matters

The one constant that I have observed in my professional life is that people underestimate the need to move fast. Of course, doing good work takes time. I once spent six months writing a URL parser. But the fact that it took so long is not a feature, it is…

State of AI: An Empirical 100T Token Study with OpenRouter https://openrouter.ai/state-of-ai

OpenRouter

State of AI | OpenRouter

An empirical study analyzing over 100 trillion tokens of real-world LLM interactions across tasks, geographies, and time.

Daniel Lemire's blog
We see something that works, and then we understand it

Telegraph | source

Telegraph

We see something that works, and then we understand it

“We see something that works, and then we understand it.” (Thomas Dullien) It is a deeper insight than it seems. Young people spend years in school learning the reverse: understanding happens before progress. That is the linear theory of innovation. So Isaac…

Daniel Lemire's blog
Why don’t we get more scientific breakthroughs?

Telegraph | source

Telegraph

Why don’t we get more scientific breakthroughs?

It is absolutely clear to me that large language models represent the most significant scientific breakthrough of the past fifty years. The nature of that breakthrough has far reaching implications for what is happening in science today. And I believe that…

Daniel Lemire's blog
Antifragile Programming and Why AI Won’t Steal Your Job

Telegraph | source

Telegraph

Antifragile Programming and Why AI Won’t Steal Your Job

Whenever I say I dislike debugging and organize my programming habits around avoiding it, there is always pushback: “You must not use a good debugger.” To summarize my view: I want my software to be antifragile (credit to Nassim Taleb for the concept). The…

https://sites.google.com/view/dpc4-2026/
4th Data Prefetching Championship 2026

Google

Home

The organizers of the 4th Data Prefetching Championship (DPC4) invite you to participate in the data prefetching competition and the subsequent workshop, which will be hosted in conjunction with HPCA 2026 at Sydney, Australia.
Summary
The goal of the competition…

属于CYY自己的世界
ROCm 7.1 + Pytorch + Strix Halo (AI MAX 395+) 踩坑小记

Telegraph | source
(author: Yangyu Chen)

Telegraph

ROCm 7.1 + Pytorch + Strix Halo (AI MAX 395+) 踩坑小记

背景 最近打算尝试微调 LLM 做一些科研有关的工作，但毕竟我之前不是做 AI 相关方向的因此自己并没有合适的 GPU 。目前多数开箱即用的模型训练都工作在 FP16 这样的精度（虽然有很多很多低精度训练的研究了），而不是推理常见的 INT4，因此仅仅 7B 参数完整载入就需要 14G 的 VRAM，再加上微调模型中所需的数据又需要额外的 VRAM，手里实在没有 VRAM 能超过 16G 的显卡。 于是，我就想起了手里 AMD Strix Halo，恰好最近（2025年11月） ROCm 7.1 刚刚发布，首次官方支持了…

Migrating the Main Zig Repository from GitHub to Codeberg https://ziglang.org/news/migrating-from-github-to-codeberg/

Chips and Cheese
Evaluating Uniform Memory Access Mode on AMD's Turin ft. Verda (formerly DataCrunch.io)
#ChipAndCheese

Telegraph | source
(author: Chester Lam)

Telegraph

Evaluating Uniform Memory Access Mode on AMD's Turin ft. Ver…

NUMA, or Non-Uniform Memory Access, lets hardware expose affinity between cores and memory controllers to software. NUMA nodes traditionally aligned with socket boundaries, but modern server chips can subdivide a socket into multiple NUMA nodes. It’s a reflection…

放个佛系aff
https://www.netcup.com/en/deals/black-friday?ref=90063

Harry Chen’s Blog
用于提供现代 PyPI 镜像的 NGINX 配置

Telegraph | source
(author: Shengqi Chen ([email protected]))

Telegraph

用于提供现代 PyPI 镜像的 NGINX 配置

众所周知，在镜像站界，PyPI 是个难伺候的主：大量的硬盘占用、巨大的流量、频繁的更新，还有不靠谱的同步工具 bandersnatch。你说为什么不靠谱？听说过其他 没有能力 删除上游删掉了的文件的同步工具吗？ 好在，伟大的 taoky 在去年写出了一个科学的同步工具 shadowmire，其中一大重要功能，就是能移除上游已经消失的版本和相关文件，为广大教育网镜像站节省了不少硬盘空间，也缓解了供应链攻击的风险。 然而，提供用户可用的 PyPI 镜像依旧没有那么简单。任何 HTTP 站点要被视作有效的 Python…

Zig Builds Are Getting Faster – Mitchell Hashimoto
https://mitchellh.com/writing/zig-builds-getting-faster

Chips and Cheese
SC25: HACCing over 500 Petaflops on Frontier
#ChipAndCheese

Telegraph | source
(author: George Cozma)

Telegraph

SC25: HACCing over 500 Petaflops on Frontier

Hello you fine Internet folks, Here at Supercomputing the Gordon Bell Prize is announced every year. The Gordon Bell Prize is awarded every year to recognize outstanding achievement in high-performance computing applications. One of this year’s finalist is…

Chips and Cheese
Qualcomm’s Snapdragon X2 Elite
#ChipAndCheese

Telegraph | source
(author: George Cozma)

Telegraph

Qualcomm’s Snapdragon X2 Elite

Hello you fine Internet folks, Last week I was in San Diego at Qualcomm’s Headquarters where Qualcomm disclosed more information about their upcoming Snapdragon X2 Elite SOC. Snapdragon X2 Elite is Qualcomm’s newest SOC for the Windows on ARM ecosystem that…

Harry Chen’s Blog
在 Linux 上对 CPU 功耗进行带内测量与限制

Telegraph | source
(author: Shengqi Chen ([email protected]))

Telegraph

在 Linux 上对 CPU 功耗进行带内测量与限制

在高性能计算场景中，能有效地测量和控制 CPU 功耗对于优化性能和能效至关重要。硬件厂商（如 Intel、AMD）提供了一些接口和工具，允许用户在不依赖外部硬件的情况下，进行带内（inband）的 CPU 功耗测量与限制。本文将简述常用的方法和工具。 基本概念 现代 CPU 的 TDP（热设计功耗）动辄上百瓦，服务器 CPU 更是达到了几百瓦的量级（如：AMD EPYC™ 9965 和 Intel® Xeon® 6962P Processor 的 TDP 都高达 500W）。再加上 turbo boost（睿频）等技术的加持，实际的…