AVX512 intrinsics for JDK’s Arrays.sort methods(github.com)

147 points by mfiguiere 3 years ago | 130 comments

Replit permanently moves to paid hosting after 7 years of free service(noreplit.com)

114 points by hackermondev 3 years ago | 79 comments

Anthropic blocks third-party use of Claude Code subscriptions(github.com)

514 points by sergiotapia 9 months ago | 426 comments

GitHub issue on the opencode repo reporting that Claude Max credentials now fail with an error saying they are only authorized for use with Claude Code.

GPU Prefix Sums: A nearly complete collection(github.com)

70 points by coffeeaddict1 1 year ago | 16 comments

Llama.cpp: Full CUDA GPU Acceleration(github.com)

640 points by gzer0 3 years ago | 265 comments

A draft GitHub pull request to llama.cpp adding full CUDA GPU acceleration for all ggml tensors, including storing the KV cache in VRAM.

New exponent functions that make SiLU and SoftMax 2x faster, at full accuracy(github.com)

368 points by weinzierl 2 years ago | 72 comments

GitHub pull request to llama.cpp by jart that upstreams llamafile's vectorized expf() functions to compute SiLU and softmax faster and more accurately on CPU.

Measuring Claude 4.7's tokenizer costs(claudecodecamp.com)

678 points by aray07 6 months ago | 474 comments

Blog post measuring Claude Opus 4.7's new tokenizer, finding about 1.47x more tokens than 4.6 on real content versus the documented 1.0–1.35x, and the resulting cost impact.

100K Context Windows(anthropic.com)

853 points by samwillis 3 years ago | 364 comments

Anthropic announcement from May 11, 2023 introducing 100K token context windows for its Claude language model.

Issue: Claude Code is unusable for complex engineering tasks with Feb updates(github.com)

1252 points by StanAngeloff 6 months ago | 686 comments

GitHub issue on anthropics/claude-code reporting that Claude Code became unusable for complex engineering tasks after the February model updates.

GLM 5.2 vs. Opus(techstackups.com)

474 points by ritzaco 4 months ago | 314 comments

A Tech Stackups article comparing the open GLM-5.2 model against Claude Opus 4.8 by having both one-shot build a 3D platformer in raw WebGL.

Tokens too cheap to meter(jyn.dev)

326 points by teoruiz 1 month ago | 214 comments

Blog post on jyn.dev arguing LLM token prices are falling by orders of magnitude, so quality and access, not token cost, will limit AI use.

Claude Code can debug low-level cryptography(words.filippo.io)

446 points by Bogdanp 11 months ago | 199 comments

Filippo Valsorda's blog post on using Claude Code to debug his new Go implementation of the ML-DSA post-quantum signature algorithm.