/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Dylan Patel

@dylan522p
150 posts
2026-03-03
Debunking the false narratives around AI Datacenters. First it was that water usage is high, but it's not. This time it's about power prices. The narrative around powe pricing is silly and politically charged with no actually backing Manipulated statistics by people with agendas
2026-03-03 View on X
Bloomberg

A UK trial by the National Grid, Nvidia, and others finds AI data centers can operate without using peak power continuously, adjusting consumption when asked

2026-02-09
Open models show 2.5x faster, 6x more expensive Lower batch size, speculative decoding harder Pareto optimal curve for Deepseek at https://inferencemax.ai/ shows this Claude Opus 4.6 is 100 Tok/s/user Deepseek at 100 is 6k Tok/s/GPU At 250 tok/s/user it's closer to 1k [image]
2026-02-09 View on X
Simon Willison's Weblog

Anthropic rolls out a fast mode for Claude Opus 4.6 in research preview, saying it offers the same model quality at 2.5x faster but costs 6x more

excited that we're making it available outside Anthropic too.Dean W. Ball /@deanwball:I would like to know more about the experimental Claude scaffold that caused Opus 4.6 to more ...

Open models show 2.5x faster, 6x more expensive Lower batch size, speculative decoding harder Pareto optimal curve for Deepseek at https://inferencemax.ai/ shows this Claude Opus 4.6 is 100 Tok/s/user Deepseek at 100 is 6k Tok/s/GPU At 250 tok/s/user it's closer to 1k [image]
2026-02-09 View on X
Financial Times

How Anthropic's bet on enterprise users is paying off; sources say Anthropic's investor guidance claims annualized revenue will exceed $30B by the end of 2026

Start-up's bet on enterprise users is paying off as coding tools fuel revenue and investor frenzy

2026-02-08
Open models show 2.5x faster, 6x more expensive Lower batch size, speculative decoding harder Pareto optimal curve for Deepseek at https://inferencemax.ai/ shows this Claude Opus 4.6 is 100 Tok/s/user Deepseek at 100 is 6k Tok/s/GPU At 250 tok/s/user it's closer to 1k [image]
2026-02-08 View on X
Simon Willison's Weblog

Anthropic rolls out a fast mode for Claude Opus 4.6 in research preview, saying it offers the same model quality 2.5 times faster but costs six times more

Opus is usually $5/million input and $25/million output.  The new fast mode is $30/million input and $150/million output!

2026-02-07
4% of GitHub public commits are being authored by Claude Code right now. At the current trajectory, we believe that Claude Code will be 20%+ of all daily commits by the end of 2026. While you blinked, AI consumed all of software development. Read more 👇 https://newsletter.semianalysis.com/ ... [image]
2026-02-07 View on X
SemiAnalysis

Analysis: Claude Code currently authors 4% of all public GitHub commits and is on track to cross 20% of all daily commits by the end of 2026

2026-02-06
4% of GitHub public commits are being authored by Claude Code right now. At the current trajectory, we believe that Claude Code will be 20%+ of all daily commits by the end of 2026. While you blinked, AI consumed all of software development. Read more 👇 https://newsletter.semianalysis.com/ ... [image]
2026-02-06 View on X
SemiAnalysis

Analysis: Claude Code currently authors 4% of all public GitHub commits and is on track to cross 20% of all daily commits by the end of 2026

2026-02-03
Google now owns more than 10% of Anthropic and XAI Google owned 14% of SpaceX https://www.spacex.com/...
2026-02-03 View on X
Bloomberg

SpaceX acquires xAI; sources say the all-stock deal values the combined company at $1.25T, with SpaceX valued at $1T and xAI at $250B; xAI burns ~$1B per month

Elon Musk's SpaceX has acquired xAI, in a deal that encompasses the billionaire's increasingly costly ambitions to dominate artificial intelligence and space exploration.

2026-02-02
Google now owns more than 10% of Anthropic and XAI Google owned 14% of SpaceX https://www.spacex.com/...
2026-02-02 View on X
Bloomberg

SpaceX acquires xAI; sources say the combined company is expected to price shares at ~$527 each in a potential IPO and would have a valuation of $1.25T

Elon Musk plans to merge SpaceX with xAI, according to people familiar with the matter, in a deal that encompasses the billionaire's …

2026-01-21
We are all calling this startup Human Sand. Sorry I make the rules, it's a fire name.
2026-01-21 View on X
New York Times

Humans&, founded by ex-Anthropic, xAI, and Google staff to build collaborative AI, raised a $480M seed from Nvidia, Jeff Bezos, and others at a $4.48B valuation

2026-01-20
We are all calling this startup Human Sand. Sorry I make the rules, it's a fire name.
2026-01-20 View on X
New York Times

Humans&, founded by ex-Anthropic, xAI, and Google staff to build collaborative AI, raised a $480M seed from Nvidia, Jeff Bezos, and others at a $4.48B valuation

As a research scientist at Anthropic, one of the world's leading A.I. companies, Andi Peng was part of a team that tried …

2025-12-27
Damn even Nvidia is tax loss harvesting Tis the season I guess
2025-12-27 View on X
Spyglass

The Groq deal secures key talent for Nvidia, including CEO Jonathan Ross, creator of the TPU, and keeps them from companies that may try to make their own chips

Twas the night before Christmas and all through the house, not a creature was stirring, not even a... wait.  What's that?

2025-12-26
Damn even Nvidia is tax loss harvesting Tis the season I guess
2025-12-26 View on X
The Information

Sources: Nvidia's licensing deal with Groq, which has raised ~$1.8B, includes payouts to Groq's key execs and investors, including BlackRock and Tiger Global

Nvidia stunned Silicon Valley on Wednesday by agreeing to pay about $20 billion to license technology from Groq …

Damn even Nvidia is tax loss harvesting Tis the season I guess
2025-12-26 View on X
Bloomberg

Nvidia agrees to a licensing deal with Groq; CEO Jonathan Ross and other top executives will join Nvidia; Groq says it will continue operating independently

anybody who knows anything about AI chips knows groq's first gen chips aren't competitive but maybe their upcoming chips might have been able to threaten nvidia? @josephjacks_ : Ov...

2025-12-25
Damn even Nvidia is tax loss harvesting Tis the season I guess
2025-12-25 View on X
Bloomberg

Nvidia agrees to a licensing deal with Groq; CEO Jonathan Ross and other top executives will join Nvidia; Groq says it will continue operating independently

Nvidia Corp. agreed to a licensing deal with artificial intelligence startup Groq, furthering its investments in companies connected …

2025-12-07
Only NVIDIA, AWS, Google have successfully deployed rack scale architecture Trn3 is the 2nd after Nvidia with switched scale up topology, which is better for frontier mixture of experts models AWS Trainium3 course corrects on software Switch design choices are... interesting
2025-12-07 View on X
SemiAnalysis

A technical deep dive into Amazon's Trainium3 accelerator, including its server SKUs' specifications, silicon design, power budget, and bill of materials

Step-Function Software & System Improvements, “Amazon Basics” GB200 NVL36x2, NL72x2/NL32x2 Scale Up Rack Architecture, Optimized Perf per TCO, Trainium4

2025-12-01
OpenAI hasn't even deployed TPUs yet and they've already saved ~30% on their entire lab wide NVIDIA fleet. This demonstrates how the perf per TCO advantage of TPUs is so strong that you already get the gains from adopting TPUs even before turning one on. The piece covers a lot [image]
2025-12-01 View on X
SemiAnalysis

An in-depth look at TPUv7 Ironwood, and how the latest Google TPU generation positions Google as the most threatening challenger to Nvidia's AI chip dominance

Fascinating article.  They argue that the reason for NVIDIA's circular investment deals is to intertwine their own fate with that of the big labs, to keep themselves on top  —  Ope...

2025-11-13
It was absolutely amazing and surreal to get to grill Satya on his AI strategy. He has some really good responses, but also some areas where we disagree. See the interview below and the report we wrote up and posted today! https://newsletter.semianalysis.com/ ...
2025-11-13 View on X
Dwarkesh Podcast

Q&A with Satya Nadella on business models for AGI, Copilot, Microsoft AI, the hyperscale business, the OpenAI partnership, capex, sovereign AI efforts, and more

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.

It was absolutely amazing and surreal to get to grill Satya on his AI strategy. He has some really good responses, but also some areas where we disagree. See the interview below and the report we wrote up and posted today! https://newsletter.semianalysis.com/ ...
2025-11-13 View on X
SemiAnalysis

A deep dive into Microsoft's AI strategy: its OpenAI deal, data center investments, neocloud renting, GitHub Copilot, its MAI models, its Maia chip, and more

SemiAnalysis :

2025-10-31
So many people hated on us for the AWS acceleration thesis a couple months ago And we were right of course Only we can do these calls due to our deep knowledge and inights with core research, datacenter, accelerator, and Tokenomics models. Amazon $AMZN Sales@semianalysis.com [image]
2025-10-31 View on X
CNBC

Amazon reports Q3 AWS revenue up 20% YoY to $33B, vs. $32.42B est., and AWS operating income up 9.5% YoY to $11.4B

Amazon said Thursday that revenue in its cloud computing business increased 20% in the third quarter, exceeding analysts' estimates.  —  Revenue at Amazon Web Services for the thir...

2025-10-29
Jensen spent 2 min on InferenceMAX at the GTC keynote 😎 InferenceMAX is completely open source and completely impartial, with the support of AMD, OpenAI, Nvidia, Microsoft, Oracle, Pytorch, SGLang, vLLM, Cursoe, Nebius, Tensorwave, etc [video]
2025-10-29 View on X
Wall Street Journal

During Nvidia's first ever GTC conference in Washington, DC, Jensen Huang praised Trump's policies and emphasized themes of patriotism and US national interests

Wall Street Journal :