HNHacker News
TopNewBestAskShowJobs

charles_irl

371 karma · joined August 6, 2024

Building useful technology out of large neural networks. https://modal.com
submissionscomments

Quail: Speed up AI-SQL by jointly optimizing query planner and inference engine

modal.com·3 pts·charles_irl·
0

>1B tokens/minute/GPU by combining query planner and inference engine

modal.com·7 pts·charles_irl·
2

How to serve trillions of tokens for trillion-parameter coding agents

modal.com·2 pts·charles_irl·
0

Routing for serverless servers with Pingora, Envoy, and Spanner

modal.com·11 pts·charles_irl·
0

Speculation Is All You Need

modal.com·12 pts·charles_irl·
0

Cutting inference cold starts by 40x with LP, FUSE, C/R, and CUDA-checkpoint

modal.com·91 pts·charles_irl·
18

How to Achieve Serverless GPUs

modal.com·8 pts·charles_irl·
0

Three types of LLM workloads and how to serve them

modal.com·75 pts·charles_irl·
5

Host overhead is killing your inference efficiency

modal.com·3 pts·charles_irl·
0

Quantized Float Exposed

quant.exposed·2 pts·charles_irl·
1

Against SQL (2021)

scattered-thoughts.net·82 pts·charles_irl·
77

Length-extension attacks are still a thing

00f.net·2 pts·charles_irl·
1

The future of Python web services looks GIL-free

blog.baro.dev·3 pts·charles_irl·
0

Lexical differential highlighting instead of syntax highlighting

wordsandbuttons.online·2 pts·charles_irl·
0

CReact – JSX for the Cloud

github.com·1 pts·charles_irl·
0

QUIC and the end of TCP sockets

codemia.io·62 pts·charles_irl·
82

In C++ modules globally unique module names seem to be unavoidable

nibblestew.blogspot.com·2 pts·charles_irl·
0

Stupid jj Tricks

andre.arko.net·3 pts·charles_irl·
0

We reverse-engineered Flash Attention 4

modal.com·5 pts·charles_irl·
0

A Tour of eBPF in the Linux Kernel: Observability, Security and Networking

lucavall.in·2 pts·charles_irl·
0

Categorical Foundations for Cute Layouts

research.colfax-intl.com·39 pts·charles_irl·
6

Pocket Casts, You Altered the Deal, So I Will Alter Your App

blog.matthewbrunelle.com·12 pts·charles_irl·
3

Modal Notebooks: How we built a cloud GPU notebook that boots in seconds

modal.com·4 pts·charles_irl·
0

Public static void main(String[] args) is dead

mccue.dev·210 pts·charles_irl·
203

Why Are Event-Driven Systems Hard?

newsletter.scalablethread.com·4 pts·charles_irl·
2

Safe C++ proposal is not being continued

sibellavia.lol·184 pts·charles_irl·
223

The unreasonable effectiveness of modern sort algorithms

github.com·6 pts·charles_irl·
0

Analyzing the memory ordering models of the Apple M1

sciencedirect.com·131 pts·charles_irl·
69

Generating diffusion QR codes that work

modal.com·5 pts·charles_irl·
2