1,076 results for out perform · 2.689s

News for “out perform”
16 results • 2686 ms server time
Moozonian News
github.com• Jun 8, 2026• 1 min read
Show HN: Mach – A compiled systems language looking for contributionsHi HN,I'm the creator of Mach (https://github.com/octalide/mach or https://machlang.org). Two days ago, we finally achieved full self hosting. I wanted to make a post here to show off the language since this is a big milestone for us.## TL;DR about the language for those curious:- There are no external dependencies anywhere in the pipeline. This includes LLVM, libc bindings, or anything of the sort (save for the historical bootstrap compiler, which requires any C compiler and has been phased out completely).- Mach is extremely opinionated and very anti-magic. WYSIWYG is a core principal for the language. There are no hidden behaviors, implicit type conversions, or "automatic features". Simplicity and stripping away ambiguity are core principals that this language upholds.- Performance currently lags behind C by about a factor of only 4x at the time of writing, almost entirely due to the lack of deep compiler optimizations like autovectorization, which have not been fully implemented ye
Moozonian News
ttanv.github.io• Jun 8, 2026• 1 min read
Show HN: Levi – run AlphaEvolve on your Claude Code/Codex for dirt cheapHi HN,Wanted to share something I'm excited about.I’ve been fascinated by AlphaEvolve and its results for more than a year now, but using open source frameworks seems overwhelming because of the high costs. I can’t really afford hundreds of Claude Opus calls every time I want to run it. I want to be able to try it out many times and all sorts of unique domains. What if it was possible for AlphaEvolve to be much more affordable while getting a better performance? .Over the last six months or so, I’ve been working on LEVI, an open source AlphaEvolve-like system that can outperform existing open source frameworks at a fraction of the cost (upto 35x cheaper!). It can also run on Claude Code or Codex, making it even more accessible (I've mostly been using it with a QWEN-30B). LEVI comes in two flavors where I felt it’ll make the most difference: Code Optimization, and Prompt Optimization (sorry math, you got a less direct path; workable through the code route).The core thesis behind LEVI is
Advertisement
Moozonian News
github.com• Jun 5, 2026• 1 min read
Show HN: Micron: a high performance C++23 (re)implementation of Libc and the STLHowdy HN,I've been working on micron for quite some time now, and I've finally gotten it to a state where I'm ready to share it with other people. I've really been pruning and testing it quite heavily for the past few weeks looking to flush out any bugs/inconsistencies. micron is a header-only C++23 core systems library that runs on Linux without libc, without libm, without the STL, and without any external dependencies at all. Fully freestanding. To my knowledge this is probably the only project of this kind, and the only libc implementation not in C. There _might_ be some Rust libc variant out there but I couldn't find any.Why does this exist? The STL makes a lot of decisions I disagree with; favoring generality over performance, safety theater that doesn't actually help anyone, ABI stability over bringing novel features to the language (regex, filesystem && company), and generally either being outdated compared to standard libraries of other languages or shipping solutions that are
Moozonian News
giovannigatti.github.io• Jun 5, 2026• 1 min read
Show HN: I benchmarked LLM agents on fixing real-world security vulnerabilitiesI built a benchmark with 20 real CVEs across 18 Python projects (Pillow, GitPython, yt-dlp, urllib3, etc). I've run it over 5 LLM agents (3 OpenAI, 2 poolside) and 3 different prompts (full advisory, locate, diagnose) with a total of 300 runs. The agents are tasked to fix security vulnerabilities in a sandboxed environment and they are scored against a hidden security tests from the maintainer's own fix.Best solve rate was 50%. On the other 50%, some fixes are sometimes coherent and pass all regression tests, but vulnerability still present.The main differentiator I found between models is cost: gpt-5.5 at 12× more expensive than gpt-5.4-mini while producing statistically similar results. Within-family performance gaps are small, which points out the difference is likely due to model training data. I also did a power analysis and the task count needed to detect a meaningful within-family edge at ~700.Full write-up: https://giovannigatti.github.io/cve-benchCode: https://github.com/Giova
Moozonian News
news.ycombinator.com• Jun 4, 2026• 1 min read
Ask HN: High school student – is learning programming still worthwhile?As a high school student, I’m trying to figure out what major I’m interested in. About half a year ago, I thought EECS was a great major for some STEM students like me, because I see many of the world's most influential entrepreneurs, such as Elon Musk and Jensen Huang, have built companies around software, hardware, and artificial intelligence, helping advance technological and society.Recently, however, I have been exposed to a variety of AI-powered programming tools, such as Claude Code and Codex. I was amazed by how capable these tools have become. They can generate websites, make software, even help solve hardware problems, and they can also help people with little programming experience build applications. On social media, I have already seen some people who have no coding background, but they use AI tools to create products and make money.This experience has led me to question whether learning programming is still worthwhile. If AI can perform many coding tasks, why should peopl
Moozonian News
fx.leftium.com• Jun 4, 2026• 1 min read
Show HN: Realistic procedural fire effect via naive algorithmSource code: https://github.com/Leftium/fx/blob/main/src/routes/fire-plas...I made this naive fire effect as realistic as possible; arguably more realistic than some simulations based on real fire fluid dynamics[1]. The naive algorithm uses a kernel to simulate convection, diffusion, and cooling of heat. Described in more detail by Joakim Hårsman[2]. Also see working demo of the OG fire algorithm by Justin Greisiger Frost[3]While canvas + CPU gets surprisingly good performance, this effect would probably run much faster on the GPU. On mobile, this version is sized to work better: https://fx.leftium.com/fire-plasma?standardSize=0&text=LeIn my improved version, flames "lick" the bottom of the floating text. This is my favorite part. It was totally unexpected and just emerged from the algorithm. Compare to Joakim's version by changing the text: https://fx.leftium.com/fire-plasma?text=%E2%97%AFMajor improvements over prior art:- True-color fire gradient/palette with transparency looks dece
Advertisement
Moozonian News
news.ycombinator.com• Jun 4, 2026• 1 min read
What are you using for distributed block storage on NVMe in production?I work on block storage at petabyte scale across AWS, Azure, and GCP. Lately I've been looking at what people actually run in production when they need shared/replicated block storage on NVMe hardware especially outside the hyperscaler managed offerings (EBS, Azure Disk, etc.). The landscape seems surprisingly thin:Ceph RBD is the default but was designed for HDDs. Most benchmarks I've seen show 15–25% flash utilization and painful tail latencies on NVMe. Crimson/SeaStore is promising but still tech preview. LINSTOR/DRBD is solid and battle tested but not NVMe-oF native. Mayastor(OpenEBS) is SPDK-based and interesting but K8s-only, one pool per node, and I've seen users report 1/4 to 1/5 of raw NVMe performance. Vitastor has impressive performance but restrictive licensing (VNPL) and bus factor of 1. Lightbits / Simplyblock are proprietary.Curious what others are running, especially:If you're at a GPU neocloud or running large AI training clusters, what's your storage stack for checkpo
Moozonian News
news.ycombinator.com• Jun 3, 2026• 1 min read
Claude Code vs. CodexI have been a long time user of Claude Code. I never felt the need to try other coding tools, and the times I did, I always came back. Across many projects, personal and professional, Claude has been my pair programmer. I have enough experience across generations of Claude models that, just by looking at the output, I can tell when a model switch has happened, strictly in the context of coding tasks. I have noticed models lean on some aspects heavily like bullet points or emojis or markdown tables or ascii diagrams or just a walls of text (4.8). Following the Claude Code 4.7 and 4.8 releases, the performance is immensely worse. Here is how I qualify "worse":1. Coding tasks: they don't follow instructions at all. They consistently miss close to a quarter of the ask, which cascades into bloat over time. 2. Reviewing existing code: they literally fail to read the code or use the tools properly.Both of these have, in my opinion, eroded the time saving value of the tool. Hard opinion: real
Moozonian News
github.com• Jun 3, 2026• 1 min read
Show HN: Agent-browser-shield – free extension to protect AI agents on the webI've been experimenting with Claude Code, ChatGPT Agent, and OpenClaw to perform more open-ended tasks for me online. A big blocker I've hit on shopping and research tasks is the agent getting a key piece of info wrong.For example, in one case, my agent decided to add a brand I don't like to the cart because the site flagged it as almost sold outThe HN crowd is probably pretty aware of the threats and can avoid them while browsing. But what about their agents?I tried prompting, but it was ineffective, because once the AI saw the threat, it polluted/distracted its contextLooking at the research, I came across a couple of papers, SusBench and Decepticon. The Deception research benchmarks indicate that increased reasoning can perform worse, because the model rationalizes the dark patternSo it seems the best approach has to be removing the information before it can pollute/poison the contextIn my day job, we have a browser extension that started as a productivity extension. However, contac
Moozonian News
lmaomoba.com• Jun 2, 2026• 1 min read
Show HN: LMAO – Temu League of Legends, Built with Opus 4.8Fun weekend project just to test out 4.8 against a pretty vanilla setup.Started out with a simple prompt, "build a temu league of legends, web-only with online, room-based multiplayer".It built it... fully functional in one shot! Then I went kinda ham and started iterating on every aspect little by little. I made it spin up subagents to each focus on designing a character. Spun up subagents to focus on designing abilities and their SFX/VFX. Did a few passes over the map, mobs, minions, etc.Then I put Ultracode Workflows to task and, across a handful of prompts, had it optimize performance, balance, and a variety of other things. Workflows is insane btw...I also used /goal pretty heavily as well. I'd sit and play and make note of like 10-15 tweaks, bug fixes, etc at a time and just let it go to town.4.8 is a one shot machine... 5.5 ain't doin this...Anyways, figured I'd share. Happy to answer any questions in the comments!Try it out solo (with bots) or with friends! https://lmaomoba.com
Moozonian News
bing.com• May 30, 2026• 1 min read
Vanguard Value ETFLearn everything you need to know about Vanguard Value ETF (VTV) and how it ranks compared to other funds. Research performance, expense ratio, holdings, and volatility to see if it's the right fund ...