ML Engineering Manager @AnthropicAI; Committer of @Scalding; Alum of (@Stripe, @Lyft, @Twitter, @Dartmouth); Forza Milan! Opinions are my own.
Bay Area, CA
Joined June 2009
- Tweets3.4K
- Following611
- Followers659
- Likes1.9K
Mu (Kevin) Lin retweeted
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.
They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
Mu (Kevin) Lin retweeted
Go try out Opus 5.5, which we launched today! It leads in coding and knowledge work, and it’s a great writer. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
Mu (Kevin) Lin retweeted
Dario is right
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: darioamodei.com/post/we-must…
Mu (Kevin) Lin retweeted
QINWEN HAS DONE IT AGAIN‼️
BACK-TO-BACK 7-5 SETS FROM 0-5 DOWN 🫨
ALT Zheng 7-5 from 0-5 vs Swiatek
Mu (Kevin) Lin retweeted
SHE'S A MIRACLE.
Qualifier Zheng Qinwen comes from 0-5 down in the decider to defeat Keys 1-6, 7-6, 7-5!
Mu (Kevin) Lin retweeted
Genuine question: How can companies like Axiom Math and Harmonic be valued at $1.5B when you can just do this with Claude? I always felt their valuation was incredibly high, but I don't see how they could ever compete with this.
Checking that a major mathematical proof is correct can take years. Formalization—converting the mathematical reasoning into a form computer proof assistants like Lean can verify—can help.
Last month, Claude completed the first formalized proof of Fermat’s Last Theorem, one of the most famous theorems of all time. This was a project experts thought would take many years. It is the largest Lean proof ever written.
Fermat’s Last Theorem was first proven in 1995 by Sir Andrew Wiles, more than 350 years after it was conjectured. Our proof, which totals over 13 million lines of code, provides machine verification. More importantly, it proves over 29,000 other theorems that the proof requires, across many areas of math which had never before been formalized.
We see this as a major step in the long process of firming up the core of mathematical knowledge, building on work from three centuries of mathematicians and hundreds of contributors to Lean and Mathlib. We are optimistic that AI-assisted verification of mathematical proofs will help reduce the burden of refereeing mathematics in an era where more proofs are being produced than ever before.
You can read about the process on our Science Blog: anthropic.com/research/forma…
And see the complete proof on GitHub: github.com/anthropics/fermat…
Mu (Kevin) Lin retweeted
Bullish on this direction. Having a metric for explanation quality makes it possible to hillclimb, and counterfactual simulatability seems right. Adam et al. created a dataset+pipeline that creates more diverse+realistic test cases than prior work & do interesting exps on it. Can also train models to write better post-hoc explanations of their behavior, as highlighted in this thread.
Mu (Kevin) Lin retweeted
Architect Labs is excited to be a part of this! AI & mathematics will have a large impact on all engineering disciplines, including chip design. We're looking forward to seeing what open problems get solved 👀
We partnered with @AnthropicAI @OpenAI
to host the world's first math hackathon.
40 hours, $2M compute - solve, understand, and present an open problem.
Applications are live: mathathonchallenge.com
Mu (Kevin) Lin retweeted
it's accelerating 🚀
better models → better kernels → better hardware utilization → faster/cheaper inference → better models
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…
Mu (Kevin) Lin retweeted
Crazy
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…
Mu (Kevin) Lin retweeted
Incredible work coming out of Architect Labs! 2 weeks is absolutely unprecedented. The world’s first AI chip designed by AI, it’s the way of the future! 🚀
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…
Mu (Kevin) Lin retweeted
we're extending our approach all the way to gds, and intend to tape it out on TSMC. more details soon 👀
Mu (Kevin) Lin retweeted
Redwood designed and verified in 2 weeks! Check it out if you haven't
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…
Mu (Kevin) Lin retweeted
Chip design × LLM in a real recursive loop — better AI, better chips.
Introducing Project Redwood 🚀🚀
@architectlabs is a frontier AI lab bringing together talent from Anthropic, xAI, Google DeepMind, and seasoned leaders across the hardware industry. We've raised a $24M seed round to build AI systems for chip design.
我們 Architect Labs 是一個 frontier AI lab ，由各家 Anthropic, xAI, Google DeepMind 還有各種硬體專家們組成，我們在做的是 ai system for chip design，目前募資 seed round $24M
-
Today, every major hardware company has its own chip design workflow—but these organizations and processes have become so large and entrenched that truly revolutionary change is difficult. We’re starting from first principles to create an AI-native chip design workflow—one that enables chip development to finally move at the speed of AI.
現在每家大硬體巨頭都有自己的晶片設計流程，但很多都大到不能做革命性的流程改動；我們在做的是從零思考，創造 ai native 的晶片設計流程，讓晶片設計真正跟上 AI 的速度。
-
Project Redwood - from a single specification, our AI system designed, verified, and deployed a chip in under two weeks—delivering 3.4× better performance per watt than NVIDIA Jetson on billion-parameter models including Llama, Qwen, and Kimi. It autonomously generated the RTL, verification, firmware, drivers, and kernels, co-designing the model, software, and silicon in one optimization loop. We acknowledge that silicon is the ultimate ground-truth. We’re taking our approach all the way to GDS. We intend to tape-out multiple improved families of Redwood co-designed for various use-cases, on TSMC.
Project Redwood - 一份規格書，我們的 AI 系統在兩週內完成晶片的設計、驗證與部署。在 Llama、Qwen 和 Kimi 等模型，其 performance per watt 比 NVIDIA Jetson 高出 3.4 倍。從 RTL、驗證、韌體、驅動到核心 kernels，全部由 AI 自己寫，並在同一個 loop 中自主設計模型、軟體與晶片。當然，流片才是最終的驗證標準。因此，我們會將這套方法一路推進至 GDS，並計畫採用台積電製程，針對不同應用場景協同設計多個持續改良的 Redwood 晶片系列並完成流片。
-
Full report on Redwood architecture and its autonomous design. Follow @architectlabs on X
我們有公開 Redwood 架構及其自主設計流程，歡迎去看論文～
Mu (Kevin) Lin retweeted
Congrats to the folks at Architect Labs for hitting this milestone!
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…
Mu (Kevin) Lin retweeted
i think we just saw one of the first real hardware loops for recursive self-improvement.
architect labs says two human architects wrote a high-level spec. its ai system generated the performance model, rtl, verification, formal proofs, firmware, drivers and kernels in under two weeks, with no human intervention below the specification.
redwood nano now runs qwen on an fpga. every block reportedly reached at least 95% code and functional coverage, and the first rtl drop reached the fpga with zero bugs.
then qwen running on redwood was connected back into the same design system. it found timing improvements and kernel optimizations for the hardware running it.
human spec → ai designs chip → chip runs ai → ai improves chip → repeat
architect projects 3.4x better performance per watt than nvidia’s jetson orin nano. important caveat: that is an asic projection. redwood has not been taped out. real silicon is the test.
the model is beginning to redesign the machine beneath the model.
announcement:
architectlabs.com/blog/redwo…
paper:
architectlabs.com/redwood-pa…
We gave our AI system a spec, and in under 2 weeks, it designed, verified, and deployed a chip that beats NVIDIA.
It’s built for low-power physical AI workloads. We’re running live inference on >B+ parameter models like Llama, Qwen, and Kimi, serving at 3.4x better perf/watt than NVIDIA Jetson.
From just a specification, our AI system autonomously generated all of the RTL Design, UVM verification, formal proofs, firmware, drivers, and kernels, co-designing the model, software and silicon as one optimization loop.
Better AI can now design better chips to run AI, leading to a loop of recursive-self improvement towards our path to abundant intelligence.
Readers added context they thought people might want to know
Architect Labs did not manufacture a physical chip; the design was tested on a commercial AMD Versal FPGA board. Furthermore, the 3.4x perf/watt gain over NVIDIA Jetson is a simulation-based projection for a future 8nm ASIC, not a physical benchmark. businessinsider.com/architect-labs… architectlabs.com/architect-labs…