注册并分享邀请链接,可获得视频播放与邀请奖励。

Max Lv 的个人资料封面
Max Lv 的头像

Max Lv (@m0d8ye)

@m0d8ye
Chip Architect (2013-Present) | Bitcoin Enthusiast since 2011. Open source developer: Views are my own.
273 正在关注    18.9K 粉丝
基于开源大模型和分布式网络的 inference 服务成为可能。接下来就是 web3 inference 服务了吧。
Wow, it has happened! 30.55 tok/s on GLM-5.2 4-bit (from @Zai_org) ran by six RTX Pro 6000's across the USA scattered over WAN! I can't believe this. It was an insane build, you can read more about it on
显示更多
Google 最先提出了了 Transformer 然后又重新发现了 MoE,并且最早设计了 AI 专用芯片 TPU,为什么现在会落后于两家 startup?
0
130
348
8
转发到社区
看来 Gemini 真的完了
I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there. It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you.
显示更多
0
17
172
0
转发到社区
中国这经济数据只能用硅进碳退来解释了。
0
35
527
24
转发到社区
已上架,想要免费版本的可以加入 testflight 或者自行编译
最近发现越来越多的网站和 app 开始用 HTTP/3 了,但是UDP 流量对于跑在 user-mode 的 proxy 是个挑战,比如 Chrome 至今不支持 Socks5 UDP Associate
显示更多
准备上架 MAS 了,改成 appex 以后用户就不需要自己去授权 sysext 了:欢迎直接加入 testflight :
A 社明显低估了中国网民翻墙的技能😅
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement:
显示更多
0
29
30
0
转发到社区
As more people cancel their Claude subscriptions, I hope that helps ease the strain on Claude’s compute resources.
为什么程序员会用 cli 里的 claude 和 codex?因为会让他们有一种自己还在写代码的幻觉。
0
133
442
12
转发到社区
1T 模型跑在 8 卡系统达到 1000 tokens/sec 这不就是明牌 B300 NVL8 了吗?
🚀 1,000+ TOKENS/S ON A 1T MODEL! 🚀 We are thrilled to release Xiaomi MiMo-V2.5-Pro-UltraSpeed in collaboration with @TileRT_AI , breaking the 1,000 tokens/s output speed on a 1 Trillion parameter model for the FIRST TIME! Not wafer-scale integration like Cerebras. Not pure on-chip SRAM chips like Groq. We achieve 1,000 tps on a 1T MoE model using just a SINGLE, STANDARD 8-GPGPU NODE. Read the full technical deep dive: Want to experience the future of real-time AI? 👉 Apply for UltraSpeed now: ⏳ Limited-Time Access: Application-based · Jun 8 – Jun 23 (PDT) 💬 Chat Experience: Completely FREE for a limited time — try the blazing-fast web chat now. ⚡ UltraSpeed API: Just 3x the price for a ~10x boost in output experience. 🤝 Enterprise & Large-Scale Needs: business-mimo@xiaomi.com
显示更多
0
15
87
5
转发到社区
不论是光伏,电动车,半导体,还是未来的机器人,最重要的是够便宜。很多人不相信中国能把这些卖到全球,因为发展中国家买不起,但其实只要卖的足够便宜就行了。那么赚不到钱卖那么多有啥用呢?其实老黄在上次那个访谈里已经说过了,最重要的是中国/华为想要建立自己的生态系统和标准体系。
显示更多
搞笑的是每次财报超预期股价都会跌,华尔街现在预期了你的超预期
NVIDIA, $NVDA, EARNINGS SUMMARY: 1. Record quarterly revenue of $81.6 billion, above expectations 2. Q1 adjusted EPS of $1.87, above expectations 3. Q2 revenue guidance of $89.2 billion to $92.8 billion, above expectations 4. New $80 billion share buyback authorization 5. Increase in dividend from $0.01/share to $0.25/share 6. Total revenue growth of +1,035% over the last 3 years Once again, Nvidia has crushed just about every expectation possible. The AI Revolution is on fire.
显示更多
说个暴论,全世界范围碳基消费都萎缩了,再纠结传统的经济数据其实没有意义。
0
38
311
12
转发到社区
说一个有意思的事情,由于过去十几年里开源社区对各种翻墙软件的不断贡献,相关协议和实现早已内化到大模型里了。比如你买个阿里云国际的服务器,在本地用任何一款国产 code agent 都可以十分钟内部署好服务器和客户端。愿意折腾的话甚至可以帮你写一套定制的混淆协议。这大概就实现了当年 shadowsocks 倡导的去中心化部署和协议混淆自定义的愿景。
显示更多
0
208
2.6K
158
转发到社区