Fast-DLLM: Training-Free Acceleration of Diffusion LLM

Abstract:Diffusion-based large language models (Diffusion LLMs) have shown promise for non-autoregressive text generation with parallel decoding capabilities. However, the practical inference speed of open-sourced Diffusion LLMs often lags behind autoregressive models due to the lack of Key-Value (KV) Cache and quality degradation when decoding multiple tokens simultaneously. To bridge this gap, we introduce a novel […]
Computer Science Courses That Don’t Exist, but Should (2015)
CSCI 2100: Unlearning Object-Oriented ProgrammingDiscover how to create and use variables that aren’t inside of an object hierarchy. Learn about “functions,” which are like methods but more generally useful. Prerequisite: Any course that used the term “abstract base class.” CSCI 3300: Classical Software StudiesDiscuss and dissect historically significant products, including VisiCalc, AppleWorks, Robot Odyssey, Zork, […]
Modern Perfect Hashing
Wojciech Muła posted about modern perfect hashing for strings and I wanted to make some comments about my own implementation (that sadly never got productionized because doubling the speed compared to gperf wasn’t really that impactful in the end). First, let’s define the problem, just so we’re all on the same page; the goal is […]
React Flow, open source libraries for node-based UIs with React or Svelte

The xyflow mono repo The xyflow repository is the home of four packages: Commercial usage Are you using React Flow or Svelte Flow for a personal project? Great! No sponsorship needed, you can support us by reporting any bugs you find, sending us screenshots of your projects, and starring us on Github 🌟 Are you […]
Apple loses UK App Store monopoly case, penalty might near $2B

A landmark case in the UK concerning Apple’s App Store practices has just been decided, with a London tribunal ruling against the company in a move that could cost Apple up to $2 billion. London tribunal rules that Apple overcharged app developers for years with unfair commissions Sam Tobin writes at Reuters: Apple abused its […]
How memory maps (mmap) deliver faster file access in Go

One of the slowest things you can do in an application is making system calls. They’re slow because you do have to enter the kernel, which is quite expensive. What should you do when you need to do a lot of disk I/O but you care about performance? One solution is to use memory maps. […]
/dev/null is an ACID compliant database
August 22, 2025 at 2 a.m. / /dev/null is web scale Atomicity Operations are “all or nothing.” Anything you write to /dev/null disappears entirely. There’s no partial write problem: it’s either written (and discarded) or not written at all. ✅ Consistency The system transitions from one valid state to another. /dev/null always stays in a […]
When is it better to think without words?
Portrait of a Man with Glasses I, Francis Bacon, 1963 This essay can be read as a complement to last year’s “How to think in writing.” Thoughts die the moment they are embodied in words.—Schopenhauer In the 1940s, when the French mathematician Jacques Hadamard asked good mathematicians how they came up with solutions to hard […]
Date bug in Rust-based coreutils affects Ubuntu 25.10 automatic updates

The Ubuntu Project has announced that a bug in the Rust-based uutils version of the date command shipped with Ubuntu 25.10 broke automatic updates: Some Ubuntu 25.10 systems have been unable to automatically check for available software updates. Affected machines include cloud deployments, container images, Ubuntu Desktop and Ubuntu Server installs. The announcement includes remediation instructions […]
Zram Performance Analysis

Introduction Zram is a kernel module that utilizes a compressed virtual memory block device allowing for efficient memory management. In this document we will analyze the performance of various compression algorithms used in Zram and their impact on the system. We will also discuss the effects of different page-cluster values on the system’s latencies and […]