1.9K
TSDylan Patel, de SemiAnalysis, afirma que la narrativa de que la IA a escala está muerta "se cae por su propio peso" cuando se observan todas las empresas que están construyendo centros de datos masivos.
@tsarnick_ia
#Semianalysis
Regardez vidéos Reels sur Semianalysis de personnes du monde entier.
Regardez anonymement sans vous connecter.
Recherches Associées
18dylan patel semianalysisinterview semianalysis dylan patel ai nvidiasemianalysis dylan pateldylan semianalysissemianalysis top 10 largest ai datacenters 2026semianalysis tsmc supply chain 2026interview semianalysis ceo dylan patel nvidiasemianalysis top 10 largest ai data centers 2026dylan patel semianalysis recent news march 2026vera rubin semianalysis
Reels en Tendance
(12)
8.6K
THDylan Patel, chief analyst at SemiAnalysis, explains how China possesses the capability to outpace the U.S. in AI development, but differing priorities have held them back. They could repurpose existing infrastructure to quickly build massive AI clusters.
@theneoniche
4.0K
DEIn this talk, the founder and Chief Analyst of @semianalysis Dylan Patel, explains that Al has made execution easy and inexpensive.
Since implementation is no longer a major hurdle, the value of a project now rests entirely on the quality of the original idea.
Power is shifting to those who control access to Al tokens.
Elite entities are already securing restricted, high-end models to gain a massive competitive edge.
Patel warns that individuals must learn to leverage these tokens to generate outsized value.
Failing to adapt to this shift risks falling into a permanent economic underclass.
Drop your thoughts on this one 💬 and you know whom to share this with 🤔
Follow [@deepcurates] for daily such insights!
🎥Credit: Invest Like The Best
* FULL VIDEO TILE ON YOUTUBE *
[ Tokenomics, Claude Mythos, and Infinite Demand ]
@deepcurates
13.6K
DWWhy AI Is Breaking Every SaaS Business Model - Satya Nadella
As part of this interview, Satya Nadella gave Dylan Patel (founder of SemiAnalysis) and me an exclusive first-look at their brand-new Fairwater 2 datacenter.
We then chatted through Satya's vision for Microsoft in a world with AGI.
Look up Dwarkesh Podcast on Youtube, Apple Podcasts or Spotify to tune in.
@dwarkesh
1.2K
TEMicrosoft was supposed to lead AI. But after pulling back on data centers and slowing down, they’re losing ground to google, oracle, and even coreweave. github copilot?
In this clip from Dylan Patel on GPT-5’s Router Moment, GPUs vs TPUs, Monetization, we dig into why Microsoft is misexecuting on the biggest tech shift in decades and whether they can turn it around before openai slips away for good.
Do you think Microsoft is blowing their AI lead?
#dylanpatel #gpt5 #microsoft #openai #azure #github #githubcopilot #microsoftcopilot #coreweave #oracle #googlecloud #ai #datacenters #chips #hyperscaler #techanalysis #semianalysis
@techno.optimist.prime
57.6K
DWWhy "Wrapping AI Models" Doesn't Work Anymore - Satya Nadella
As part of this interview, Satya Nadella gave Dylan Patel (founder of SemiAnalysis) and me an exclusive first-look at their brand-new Fairwater 2 datacenter.
We then chatted through Satya's vision for Microsoft in a world with AGI.
Look up Dwarkesh Podcast on Youtube, Apple Podcasts or Spotify to tune in.
@dwarkesh
141
TWWaleed Atallah, CEO and Co-founder of Makora, talks about what AI automation is really pushing toward:
delegation.
Not just assistance.
As model capability keeps improving, the question becomes less about “can it help?” and more about “can I trust it to do this end-to-end?”
Hosted by @dylan522p Dylan Patel of @semianalysis SemiAnalysis for @read.sail SAIL Media x SemiAnalysis.
Big conversations like this are exactly why compute infrastructure is becoming one of the most important layers in AI.
At Twill, we explore the future of agents, deployment, and performance engineering.
👉 https://luma.com/twill
#AI #Agents #Compute #GPU #SemiAnalysis SAIL
@twillsocials
1.3K
NOOpenAI’s open source model pushes the commodity bar up and makes AI far easier to adopt. SemiAnalysis founder and CEO Dylan Patel shares his thoughts.
@nopriorspod
3.1K
SUSupermicro in Silicon Valley
Supermicro founder & CEO, Charles Liang sits with Dylan Patel, CEO of @SemiAnalysis for an insightful conversation on the evolution of Silicon Valley and Supermicro's early focus on server manufacturing in the USA.
Dive into how Supermicro engineered Data Center Building Block Solutions® (DCBBS): link in story
#MadeinAmerica #DCBBS #AI #Podcast
@supermicro_smci
631
THDoug O'Laughlin of SemiAnalysis makes a bold claim about @nvidia's dominance in the AI chip market.
"We can talk about all this, but there really hasn't ever been a single successful accelerator startup company. Name one."
But he tells Akash Pasricha that customers don't want a monopoly.
"I don't think anyone is really interested in having one player have 95% of the market."
Watch the full episode: https://bit.ly/3HRdGwj
@theinformation
452
TI668 Years of AI = 1 Burger 🍔💧
One cheeseburger equals 668 years of continuous AI usage. 🤯
New research from SemiAnalysis compared xAI’s Colossus 2 data center (one of the world’s largest) to In-N-Out Burger locations. The annual water footprint? Equal to just 2.5 burger joints.
But here’s where it gets complicated…
Cornell researchers say location changes EVERYTHING. A sustainable facility in one region could devastate another’s water supply.
Maybe the real question isn’t “how much water” but “whose water, and where?”
That’s a much harder conversation to have. 💧
What do you think? Drop your thoughts below. 👇
#AIWaterUsage #DataCenters #Sustainability #TechDebate
@timothyjbish
339
SEHere is a fun video I made sometime last year based on a @semianalysis article on the economic use-cases of quadropeds robots. SemiAnalysis are the top-dogs in the semiconductor / AI industry research scene. I'd highly recommend subscribing to their free newsletter for in depth coverage of the largest infrastructure buildout of all time.
Robots are a super interesting offshoot of AI development as so much of the success of robotics depends on the ability of AI models* to bridge the gap between the media that humans use to convey meaning and the world of sensors that make up the vantage-point of robots. By this I mean that the visual/textual/audio representration of our shared human experience is comparatively easy to mimic effectively because we create such extensive records of it, whereas to effectively mimic the world of a robot operating in the world needs a new source of data.
*specifically 'world models', which are early in development compared to the models we use now, because of the currently-out-of-reach economic use-cases that these models would unlock. Building an realistic-enough world model would theoretically allow a model for robot operation to be trained against this world model. The robot-operation model would be able to experiment endlessly inside this model, unbound by the physical-world restrictions that currently hold back the development of these models (ie not enough time and not enough robots to test on)
@sebastian_borckenhagen
✨ Guide de Découverte #Semianalysis
Instagram héberge thousands of publications sous #Semianalysis, créant l'un des écosystèmes visuels les plus dynamiques de la plateforme.
Découvrez le dernier contenu #Semianalysis sans vous connecter. Les reels les plus impressionnants sous ce tag, notamment de @dwarkesh, @theneoniche and @deepcurates, attirent une attention massive.
Qu'est-ce qui est tendance dans #Semianalysis ? Les vidéos Reels les plus regardées et le contenu viral sont présentés ci-dessus.
Catégories Populaires
📹 Tendances Vidéo: Découvrez les derniers Reels et vidéos virales
📈 Stratégie de Hashtag: Explorez les options de hashtags tendance pour votre contenu
🌟 Créateurs en Vedette: @dwarkesh, @theneoniche, @deepcurates et d'autres mènent la communauté
Questions Fréquentes Sur #Semianalysis
Avec Pictame, vous pouvez parcourir tous les reels et vidéos #Semianalysis sans vous connecter à Instagram. Aucun compte requis et votre activité reste privée.
Analyse de Performance
Analyse de 12 reels
💡 Posts top moyennent 20.9K vues (2.7x au-dessus moyenne)
Publiez régulièrement 3-5x/semaine aux heures actives
Conseils de Création de Contenu et Stratégie
💡 Le meilleur contenu obtient plus de 10K vues - concentrez-vous sur les 3 premières secondes
✍️ Légendes détaillées avec histoire fonctionnent bien - longueur moyenne 536 caractères
📹 Les vidéos verticales de haute qualité (9:16) fonctionnent mieux pour #Semianalysis - utilisez un bon éclairage et un son clair
✨ Beaucoup de créateurs vérifiés sont actifs (33%) - étudiez leur style de contenu
Recherches Populaires Liées à #Semianalysis
🎬Pour les Amateurs de Vidéo
Semianalysis ReelsRegarder Semianalysis Vidéos
📈Pour les Chercheurs de Stratégie
Semianalysis Hashtags TendanceMeilleurs Semianalysis Hashtags
🌟Explorer Plus
Explorer Semianalysis#dylan patel semianalysis#interview semianalysis dylan patel ai nvidia#semianalysis dylan patel#dylan semianalysis#semianalysis top 10 largest ai datacenters 2026#semianalysis tsmc supply chain 2026#interview semianalysis ceo dylan patel nvidia#semianalysis top 10 largest ai data centers 2026