SemiAnalysisMédia·en·il y a 2 jours·via RSSThe Huawei-DeepSeek Symbiosisnewsletter.semianalysis.comSemiAnalysis·il y a 2 jours·newsletter.semianalysis.comil y a 2 joursThe Huawei-DeepSeek SymbiosisHuawei Ascend 950PR Deep dive & µbenchmarks, DSv4.1 Flash Analysis, DeepSeek support for Ascend
SemiAnalysisMédia·en·il y a 15 jours·via RSSBeijing Will Not Pace the Frontier: China’s Speed-First AI Safety Regimenewsletter.semianalysis.comSemiAnalysis·il y a 15 jours·newsletter.semianalysis.comil y a 15 joursBeijing Will Not Pace the Frontier: China’s Speed-First AI Safety RegimeAI safety is on fire.
SemiAnalysisMédia·en·il y a 18 jours·via RSSAnthropic Subscriptions Offer 5x+ More Value Than OpenAInewsletter.semianalysis.comSemiAnalysis·il y a 18 jours·newsletter.semianalysis.comil y a 18 joursAnthropic Subscriptions Offer 5x+ More Value Than OpenAILimit testing every AI subscription plan from Anthropic, OpenAI, Meta, SpaceXSI, MiniMax, Moonshot, Z.ai, Cursor, and Cognition
SemiAnalysisMédia·en·il y a 25 jours·via RSSHow GLM5.3 Sparse Attention Affects HBM Memory Usagenewsletter.semianalysis.comSemiAnalysis·il y a 25 jours·newsletter.semianalysis.comil y a 25 joursHow GLM5.3 Sparse Attention Affects HBM Memory UsageGLM-5.3, KV Cache Offloading, HiSparse, DeepSeek Sparse Attention, IndexShare, Single-rollout Asynchronous Optimization, Cybersecurity
SemiAnalysisMédia·en·il y a 28 jours·via RSSThe Chinese AI Infrastructure Boom: Introducing the SemiAnalysis China Datacenter Modelnewsletter.semianalysis.comSemiAnalysis·il y a 28 jours·newsletter.semianalysis.comil y a 28 joursThe Chinese AI Infrastructure Boom: Introducing the SemiAnalysis China Datacenter Model1,000+ facilities across 60+ operators mapped, built retail-first and flipped by AI, largest hyperscaler leases 1/5 national capacity, 100MW in 12 months, Eastern Data Western Compute
SemiAnalysisMédia·en·il y a 1 mois·via RSSEngrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloadingnewsletter.semianalysis.comSemiAnalysis·il y a 1 mois·newsletter.semianalysis.comil y a 1 moisEngrams Embedding Entendre: Codesign for Efficient DRAM/SSD OffloadingNew Model Architecture Implications for TAM of DRAM/NVMe, DeepSeek V4.1 Flash, AgentX, InferenceX, NVMe experiments
SemiAnalysisMédia·en·il y a 1 mois·via RSSA Brain Too Big to Carry — On-Device vs Datacenter Inferencenewsletter.semianalysis.comSemiAnalysis·il y a 1 mois·newsletter.semianalysis.comil y a 1 moisA Brain Too Big to Carry — On-Device vs Datacenter InferenceRobot Models, Silicon Efficiency, Jetson Thor vs. B300 TCO, Deployments, The Network Wall
SemiAnalysisMédia·en·il y a 1 mois·via RSSNvidia’s Backstop Universe – Heads I Win, Tails Who Loses?newsletter.semianalysis.comSemiAnalysis·il y a 1 mois·newsletter.semianalysis.comil y a 1 moisNvidia’s Backstop Universe – Heads I Win, Tails Who Loses?The $11T AI Buildout, Nvidia’s Backstop Economics, and the Limits of Nvidia’s Balance Sheet
SemiAnalysisMédia·en·il y a 2 mois·via RSSKorea’s Trillion-Dollar Sovereign AI Investment: Nvidia Wins, Hynix Losesnewsletter.semianalysis.comSemiAnalysis·il y a 2 mois·newsletter.semianalysis.comil y a 2 moisKorea’s Trillion-Dollar Sovereign AI Investment: Nvidia Wins, Hynix LosesKorea hosts a Squid Games, National AI Tournament, the best non-Chinese open source model gets eliminated, why Nvidia needs open source, implications for Hynix and Samsung
