Search across episodes, people, companies, and topics — or ask the AI agent for deeper research and structured analysis.

Browse all available podcasts, episodes, and categories. Explore Technology shows and discover new content to follow.

Manage your subscribed podcasts. New episodes are automatically analyzed with AI when published, so you never miss important discussions.

Your AI analysis feed. Every episode you analyze appears here with key insights, summaries, and extracted intelligence.

Browse individuals mentioned across analyzed episodes. See who's being discussed, in what context, and how often.

Track companies and organizations referenced in podcast discussions. Monitor brand mentions and competitive intelligence.

Explore recurring topics and themes extracted from analyzed episodes. Discover what subjects are driving podcast conversations.

Spot emerging trends identified by AI across analyzed episodes. See what's gaining momentum in podcast discussions.

Browse brands and advertisers identified across analyzed episodes. See which companies sponsor which shows and how often.

Batch size optimization and latency-cost tradeoffs

Discussed in 1 analyzed podcast episode across 1 show

Discussed On

Dwarkesh Podcast

Episodes

Dwarkesh Podcast

Dwarkesh Podcast · Apr 29, 2026

Reiner Pope – The math behind how LLMs are trained and served

Roofline analysis for transformer inference Memory bandwidth as binding constraint on inference Mixture of experts parallelism and communication patterns Expert parallelism vs tensor parallelism vs pipeline parallelism KV cache memory requirements and amortization