IOSF (@IOSFVC) header
IOSF (@IOSFVC) profile picture

Intelligence at Frontier https://nitter.cf/t.co/BgrhvbVkdc

Joined September 2026
Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)Photo by IOSF (@IOSFVC)
Late-night talks about inference, GPUs and everything SGLang 🍸 Thank you to everyone who came to the SGLang Summit Afterparty, co-hosted with @hyperbolic_labs and SourceM3d! It was great catching up with old friends, meeting new ones, and hearing what everyone is building, from inference platforms to GPU clouds and compute marketplaces. Big congrats again to @radixark on the $100M seed. The SGLang community keeps growing. Where should we host the next one?
1
1
10
755
🧠 We went through the research on what it costs to run an AI agent. The expensive part isn’t the model. It’s the room it lives in. Meta’s personal agent reportedly costs $20/month. Like Google’s and OpenAI’s, it runs on its own cloud computer, called a sandbox. 1. The lock is free, the rent isn’t. Isolation tech is open source, at ~8¢/hr list price. But researchers priced a small sandbox across 12 providers at an est. $24–$124/month if left on. 2. Most of the rent pays for waiting. In a Microsoft study of 13.5M Copilot agent sessions, 85.4% of the time went to waiting for the model and 4.7% to running tools. 3. Shutting it down isn’t free either. Freezing takes ~4s/GB against a typical 5.8s pause. In one AWS replay, shutting down after every tool call pushed the total bill to 4.9x leaving it on. 4. The money is moving. Modal says its sandbox business runs at $300M/yr. Baseten bought Blaxel. AWS, Google and Cloudflare are bundling sandboxes. 💡 Our read: the hard problem isn’t isolation, it’s state. The winners will stop paying rent while the agent waits. 🎬 Full breakdown in the video below 👇
4
143
Come and join us!
Heading to SGLang Summit this week? Keep the conversation going after with us! We’re co-hosting From GPU to Token, an afterparty with @IOSFVC and Sourcem3d for the people building and serving models at scale: inference, kernels, serving infra, and everything between the GPU and the token. 🗓 Thu, Nov 12 · 6–9pm PT · San Francisco 🍻 Drinks and bites on us Space is limited, so RSVP for approval: luma.com/xubn83xh
78
🍽️IOSF Founder Table Vol. 1 | From Inference to RSI Great AI conversations don’t always happen on stage. Sometimes, they happen around a table. For our first gathering in San Francisco, researchers from OpenAI, xAI & Google DeepMind joined founders and investors to exchange ideas on inference economics, world models, recursive self-improvement (RSI), and AI agents. A few takeaways: 🔶 Inference: Customer retention and sustainable margins matter as much as growing demand. 🔶 World models: Impressive demos are just the beginning. Real use cases drive adoption. 🔶 RSI: Faster research doesn’t always mean better decisions. 🔶 AI agents: Greater autonomy makes oversight and trust more important. 💬 Interested in joining a future gathering? DM us or leave a comment! 🔗 Learn more about our events: iosf-gatherings.vercel.app #IOSF #FounderTable #AIResearch
3
165
🚀 From GPU to Token, join us in San Francisco! Together with @hyperbolic_labs and SourceM3D, we’re bringing the AI compute ecosystem together for an evening of conversations and connections following SGLang Summit! ⚡From data centers and GPU clouds to inference and AI applications, let’s connect the people powering the next generation of AI. Expect guests from SGLang, NVIDIA, GMI Cloud, Apple, Google, Amazon & more! 📅 Nov 12 | 6–9 PM PST 📍 San Francisco, CA 🎟️ RSVP: luma.com/xubn83xh Let’s connect, exchange ideas, and make things happen! 🤝 #AIInfrastructure #GPU #SGLang
1
5
234
IOSF is doubling down as @hyperbolic_labs announces its latest round. How we saw it: In 2024 the open question wasn't whether inference would grow. It was whether the newest GPUs would just obsolete everything that came before. Our read was they wouldn't. Inference was going to keep climbing, and older cards would slide down the cost curve into new buyers and new workloads, not get thrown away like last year's phone. The hard part wasn't owning more GPUs. It was about making heterogeneous compute supply unified, liquid, productive, and durable. That's the layer Hyperbolic built. Making messy supply easy to access, manage, and orchestrate. Two years later that call still looks right. And a new demand is showing up on top of it: hedging compute. Long GPU books want price protection, not just capacity. Indices are easy; physical settlement isn't. Hyperbolic is rare in being able to do both: take multi-year supply, break it into short-term product customers actually buy, and sit underneath the hedge as the party that can deliver the hardware. We're proud to continue our partnership with @zjasper and the @hyperbolic_labs team on turning compute into a liquid market.
We’ve raised an oversubscribed $30M Series A1 led by @walden_catalyst to build the Global Compute Grid. We’re turning fragmented GPU capacity into one unified platform where AI teams can access, manage, sell, and orchestrate compute. Our mission: make the world’s compute unified, productive and liquid.
1
2
6
394
China AI Weekly | Oct 28 – Nov 4 Where the capital moved this week: 💰 @deepseek_ai raised $15.9B, China's largest startup round ever. It alone lifted October global VC to $59B (+45% YoY); AI took 66% of all dollars. 🔌 Compute is going domestic. @deepseek_ai reportedly ordered ~160K Huawei Ascend chips (~$2.5B) for V5 (unconfirmed). Nvidia's China AI chip share is down to ~34%. 📱 @Apple Intelligence launches in China in December on @Alibaba_Qwen . $BABA +5%. 🤖 @deepseek_ai previewed a multi-agent "super agent" and shipped Creator Mode — going straight at the coding-assistant market. ⚖️ NDRC is drafting full-lifecycle AI safety standards; a financial regulator warned AI investment is "overly concentrated." 📉 Counter-signal: Chinese AI stocks are sliding, and private benchmarks show a ~12-pt gap to US frontier models. Our read: mega-rounds at the top, compression everywhere else. Distribution now beats leaderboard rank. Follow @IOSFVC for the weekly China AI → capital signal.
4
208
GPU越便宜,AI巨头越慌?🔥💸 OFR House EP07|GPU一降价,整个AI融资循环会崩? 头部实验室约40%的算力用于推理,英伟达产能排到2029年,又联手华尔街撬动$500B+算力融资;一旦新模型更强更便宜、不再需要那么多卡,显卡降价就可能引发连锁清算 #AI 本期特别转载自播客《OtherWise他山之玉》,主理人 @harrychen404 对话 @IOSFVC 首席投资人@glazecl,硅谷一线投AI的华人VC 🎙️ 聊巨头喊停AI训练背后的上市算盘、老黄和模型厂的同床异梦、YC为什么在走下坡路、小团队如何在大厂眼皮底下活下来,以及为什么说美国VC本质上是一款卖给LP的产品👀 🎧小宇宙:xiaoyuzhoufm.com/episode/6ae… 🍎Apple:podcasts.apple.com/jp/podcas… 📸YouTube:youtu.be/edCMcJE6xZo
3
1
6
550
For AI agents, price per token is the wrong number. @deepseek_ai V4.1 Flash: $0.003 per million input tokens on a cache hit vs $0.15 fresh (off-peak). That's 50×. Our 60-second video explains why this matters for agents, how DeepSeek got here, and the caveats 👇
3
120
China’s AI race is shifting from model size to data control. But who are the people actually shaping that shift? As high-quality public training data becomes harder to scale, the teams responsible for sourcing, training data, and evaluation are becoming increasingly central to AI development. A closer look at 8 major Chinese AI companies reveals the people behind this transition: • @BytedanceTalk : Yinglei Wang oversees AI Data & Safety, while Chenggang Li leads pretraining data within Seed. • @TencentGlobal : Shunyu Yao has taken on leadership across foundation models, AI infrastructure, and AI Data. • @AlibabaGroup : Minggang Wu, Zhiwei Yang, Ze Xu, and Zenan Meng cover training data, expert data, evaluation, and embodied AI data. • @meituan : Qiong Huang is reportedly responsible for training data within its LongCat model team. • @xiaohongshu : Former DeepSeek researcher Guanting Chen leads pretraining data. • @deepseek_ai : Yishi Pu and Yukun Li are key figures in its training data operations. • @moonshot AI: Yangyang Liu is a key contributor to its multimodal data and evaluation research. • @ZhipuAI AI: Xiaohan Zhang has played a central role in building its data annotation infrastructure. The bigger picture: data is no longer just a supporting resource. The people deciding what models learn, how data is produced, and how progress is measured are becoming increasingly influential. As China’s AI industry moves beyond scaling models, these data teams are worth watching. Explore the map below. 📷 Based on reporting by Z Finance. Roles reflect reported information and may change.
2
117
1/ One AI bill, 14 models. DeepSeek V4.1 Flash handled 53% of the tokens for 2.7% of the spend. $0.02 per million tokens. The bill's average: $0.41. The trick isn't the model. It's the cache. 🧵
2
1
3
165
6/ The catch: fresh prompts pay $0.15-0.30 in, $0.60-1.20 out weekday peak hours cost 2x cache hits need an exact prefix match prices rose 50% to 1,100% by rate in Aug Anthropic has twice accused DeepSeek of distilling Claude
1
14