19 March 2024

30 links · Tuesday

GoblinTools - Neil Turner's Blog

If you sometimes find doing things overwhelming, or need help converting some disparate thoughts into something coherent, then GoblinTools may help you. It was recommended to me by someone on Mastodon some time ago; sadly I’ve lost track of who tooted or boost…

Why does an extraneous build step make my Zig app 10x faster?

For the past few months, I’ve been curious about two technologies: the Zig programming language and Ethereum cryptocurrency. To learn more about both, I’ve been using Zig to write a bytecode interpreter for the Ethereum Virtual Machine. Zig is a gr…

Execution is King

As a Product Manager, more often than not, I notice people mixing up _ideas_ and _execution_ in discussions. Both these concepts have entirely unique levels of fidelity to what the finished product will be. It’s important for Product Managers to know the diffe…

Why Fastmail over ProtonMail - Scott Spence

I’ve now been using Fastmail for a year now. I was using ProtonMail previously but decided to migrate to Fastmail. I’m going to go into the details of why I switched. This is personal preference and my reasoning behind the switch. For full...

Let's create a Tree-sitter grammar

One of my favorite features in Neovim is the Tree-sitter integration. It allows for fast syntax highlighting that works well even in an error state (often the case when you’re editing code), and it has additional semantics (you can differentiate between functi…

How to Start Google

(This is a talk I gave to 14 and 15 year olds about what to do now if they might want to start a startup later. Lots of schools think they should tell students something about startups. This is what I think they should tell them.)

GitHub - microsoft/garnet: Garnet is a remote cache-store from Microsoft Research that offers strong performance (throughput and latency), scalability, storage, recovery, cluster sharding, key migration, and replication features. Garnet can work with existing Redis clients.

Garnet is a remote cache-store from Microsoft Research that offers strong performance (throughput and latency), scalability, storage, recovery, cluster sharding, key migration, and replication feat...

LLM inference speed of light

In the process of working on calm, a minimal from-scratch fast CUDA implementation of transformer-based language model inference, a critical consideration was establishing the speed of light for the inference process, and measuring the progress relative to tha…