30초 요약
Walter Goodwin discusses placement, routing and experiment latency, arguing that faster AI assistance does not make every chip-design step instantaneous.
- 1Placement and routingSome design steps remain computationally expensive.
- 2Automation and latencyAutomating thought does not remove every delay.
- 3Experiment bottlenecksTesting still consumes time and resources.
핵심 정리
- Placement and routing can take substantial compute time.
- Some reasoning can be automated without eliminating all latency.
- Experiments still need time and resources.
챕터별로 다듬은 전문 읽기
Chip-design latency
Fractile CEO Walter Goodwin discusses computationally expensive steps in chip design, including placement and routing.
Fractile CEO Walter Goodwin discusses computationally expensive steps in chip design, including placement and routing.
He says conventional layout tools can take days for a design, even if other parts of the work are assisted by AI.
Goodwin distinguishes automating some reasoning from removing the intrinsic latency of experiments or computational processes.
Experiments still take time
He draws an analogy to AI model development: better systems may guide experiments, but experiments still take resources and time.
He draws an analogy to AI model development: better systems may guide experiments, but experiments still take resources and time.
His point is that expensive design and validation steps can constrain the speed of progress, even when other tasks accelerate.
같은 채널의 다른 글
38:06EN
No Priors: AI, Machine Learning, Tech, & Startups· 2026년 10월 29일팟캐스트 · 3명Serval’s Jake Stauch on coding agents for enterprise IT게스트 Jake Stauch · 진행자 Sarah Guo, Elad GilStauch explains why Serval is challenging ServiceNow, how it approaches permissions and approvals, and why background agents remain difficult to deploy safely.
1:07ENNo Priors: AI, Machine Learning, Tech, & Startups· 2026년 10월 27일팟캐스트George Kurtz on How AI May Change Cyberattacks게스트 George KurtzThe CrowdStrike CEO describes his assessment of evolving threats.
39:23ENNo Priors: AI, Machine Learning, Tech, & Startups· 2026년 10월 15일팟캐스트 · 2명The Defenders' Window: Greg Brockman on AI-Generated Knowledge, Drug Discovery, and the New Physics of Cybersecurity진행자 Sarah Guo, Elad GilOpenAI's president explains why solving open math problems is a proof point for science, how proactive agents change product form, and why security spending must exceed software spending.
1:09:44ENNo Priors: AI, Machine Learning, Tech, & Startups· 2026년 10월 9일팟캐스트 · 2명10,000 GB300s for Four Weeks: Reflection’s Open-Model Ledger진행자 Sarah Guo, Elad GilMisha Laskin tells No Priors how Reflection trained Beam, why reinforcement learning outspent pre-training, and why the West needs its own open weights.
35:39ENNo Priors: AI, Machine Learning, Tech, & Startups· 2026년 10월 2일팟캐스트 · 2명Compute Scaled a Millionfold. Bandwidth Only 40x. Fractile Wants to Fix That Ratio진행자 Sarah Guo, Elad GilIn the summer of 2022, Walter Goodwin founded Fractile in London to build very fast inference chips for the world's largest models.
1:00ENNo Priors: AI, Machine Learning, Tech, & Startups· 2026년 9월 29일팟캐스트Michael Lee on Incentives in Services and Software게스트 Michael LeeA short critique of how business incentives shape technology adoption.