这可能是我第一次在网上出镜（

Personal AI computing has just become more accessible than ever before with the new DGX Spark from NVIDIA - a tiny yet powerful Blackwell GPU in your hand. Check out the firsthand unboxing video and blog from the SGLang community members yv and Richard Chen.

Congrats on NVIDIA DGX Spark⚡️We have a YouTube video this time given by yv 🤩

🚀 Excited to collaborate with NVIDIA and SemiAnalysis on pushing inference performance to the next level! On the Blackwell GB200 NVL72, SGLang achieved 26K input / 13K output tokens per GPU/sec. On the SemiAnalysis InferenceMAX benchmark, SGLang is the default engine for

Andrej越来越会整活了，但他也越来越独立开发者了。。

他终于g了

这也没错其实，大部分业务逻辑就是map和reduce，再加个filter，也就能规约成for和if

那这张图可以扩展一下：VSCode, Cursor, Windsurf, ChatGPT.app, Claude.app 都是Chromium

Exciting updates on DGX Spark: Now you can run gpt-oss-20b at 70 tokens/s with SGLang! This is 1.4x faster than what we got in our blog last week. We worked with the NVIDIA AI Developer team to fix a bunch of Triton and quantization issues. Cannot wait to see how much performance we

1.4x speedup after one week of release!

我两年前的这条推含金量还在上升

SGLang now runs natively on TPU with a new pure Jax backend! SGLang-Jax leverages SGLang's high-performance server architecture and uses Jax to compile the model's forward pass. By combining SGLang and Jax, it delivers fast, native TPU inference while maintaining support for

float64 is all you need

Thought experiment: Model the entirety of a brain (in a computer) and mock everything else. Do you get consciousness?

During the LF InfiniEdge AI Meetup on Oct 23, we demoed running GPT-OSS on the NVIDIA DGX Spark with SGLang, hitting blazing-fast speeds: 70 tps on 20B and 50 tps on 120B ⚡️ We also showed how to use Open WebUI and even Claude Code with the local SGLang. Here’s our recap blog

sglang支持图像和视频模型了诶

草，营销号拿1个GPU rack和144个TPU rack去比较性能真是xswl