Watched Feeds (1)
SemiAnalysis
Alert recipients
Leave blank to use the global distribution list.
Recent Articles
30 loaded
SemiAnalysis
Anthropic’s Claude 4.6 Opus and Claude Code have soared in demand. Anthropic’s ARR has nearly tripled in just a single quarter from $9B at the end of last year to over $25B today. Open models such as GLM and Kimi K2.5 caused open model use cases to soar. Capital raises by firms like Anthropic, OpenA
SemiAnalysis
Nvidia’s Datacenter Blackwell GPU (SM100) represents one of the largest GPU microarchitecture change in a generation, yet no detailed whitepaper exists. Until today, there is no public datacenter Blackwell architecture microbenchmarking study on PTX and SASS instructions, such as UMMA and TMA, with
SemiAnalysis
Source: Nvidia
At GTC 2026, Nvidia delivered an event packed full of ground breaking announcements. Nvidia’s pace of innovation is not showing any signs of slowing, as they introduced three entirely new systems this year: Groq LPX, Vera ETL256, and STX. Also announced were updates to Nvidia’s Kyber
SemiAnalysis
SemiAnalysis x Fluidstack Hackathon
We’re hosting a hackathon ahead of GTC this year on Sunday the 15th, in partnership with Fluidstack, and would love to have you join us!
Apply here: https://luma.com/SAxFSHack
The Compute Shortage
Token demand is skyrocketing and the need for AI compute continues
SemiAnalysis
Are AI Datacenters Increasing Electric Bills for American Households?
SemiAnalysis x Fluidstack is launching GTC with a 48-hour, full-stack AI infrastructure hackathon on March 15th from Power to Prefill, Dirt to Decode. With speakers from OpenAI, GPU MODE, and Thinking Machines, plus compute grants
SemiAnalysis
At CES 2026, Nvidia officially announced in detail all 6 Rubin platform products: the Rubin GPU, Vera CPU, NVLink 6 Switch, ConnectX-9, BlueField-4, and Spectrum-6. VR NVL72 is the second generation of Nvidia’s rack scale Oberon architecture that takes the stage. With competition catching up on rack
SemiAnalysis
Introduction
InferenceXv2 (formerly InferenceMAX) builds on the foundation established by InferenceMAXv1, our open-source, continuously updated inference benchmark that has set a new standard for AI inference performance and economics. InferenceMAXv1 moved beyond static, point-in-time benchmarks by
SemiAnalysis
Since 2023, the datacenter story has been simple. GPUs and networking are king. The arrival and subsequent explosion of AI Training and Inference have shifted compute demands away from the CPU. This meant that Intel, the primary supplier of server CPUs, failed to ride the wave of datacenter buildout
SemiAnalysis
Prices of memory are going crazy. SemiAnalysis has been calling this out for over a year since late 2024. The scariest thing is that we aren't even close to the peak. We go through fab by fab production and expansion versus detailed end market demand by memory type to forecast memory revenue, pricin
SemiAnalysis
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.
Our sister publication Fabricated Knowledge described soft
SemiAnalysis
Tokens and burgers are not two ideas you often see paired in the same title. If you’re wondering what they could possibly have in common, stick with us for this short piece, because today we’re putting them face to face in a duel over a hot topic - datacenter water usage. If you want a different per
SemiAnalysis
It’s an odd time in the chipmaking industry. On one hand, we are ramping into the biggest supercycle ever seen. Advanced logic, DRAM, NAND, customers can’t get enough, chipmakers can’t build capacity fast enough, and we may soon be limited by fab equipment supply. On the other hand, the rapid techno
SemiAnalysis
In 2013, TSMC made a $10 billion bet on a single customer. Morris Chang committed to building 20nm capacity with uncertain economics on the promise that Apple would fill those fabs. “I bet the company, but I didn’t think I would lose,” Chang later said. He was right. Apple’s A8 chip launched in 2014
SemiAnalysis
We’re hiring for AI Analysts and Tokenomics Analyst roles. Apply here or reach out directly.
Last June, we argued that scaling RL is the critical path to unlocking further AI capabilities. As we will show, the past several months have affirmed our thesis: major capability gains are coming from rampi
SemiAnalysis
Co-Packaged Optics (CPO) has long promised to transform datacenter connectivity, but it has taken a long time for the technology to come to market, with tangible deployment-ready products only arriving in 2025. In the meantime, pluggable transceivers have kept pace with networking requirements and r
SemiAnalysis
The Grid is Old and Tired
Nearly two years ago, we were the first to predict a looming power crunch. In our report AI Datacenter Energy Dilemma - Race for AI Datacenter Space, we forecasted AI Power Demand in the US to grow from ~3GW in 2023 to over 28GW by 2026 – a pressure that would overwhelm Ame
SemiAnalysis
Trainium3: A New Challenger Approaching!
Hot on the heels of our 10K word deep dive on TPUs, Amazon launched Trainium3 (Trn3) general availability and announced Trainium4 (Trn4) at its annual AWS re:Invent. Amazon has had the longest and broadest history of custom silicon in the datacenter. While th
SemiAnalysis
Earlier this year, U.S. President Donald Trump called TSMC “one of the most important businesses in the world.” Although he was incorrect, TSMC is the most important business in the world, his government and others do realize the importance of bringing TSMC into their countries.
TSMC CEO C.C. Wei,
SemiAnalysis
The two best models in the world, Anthropic’s Claude 4.5 Opus and Google’s Gemini 3 have the majority of their training and inference infrastructure on Google’s TPUs and Amazon’s Trainium. Now Google is selling TPUs physically to multiple firms. Is this the end of Nvidia’s dominance?
The dawn of the
SemiAnalysis
Microsoft was at the top of AI in 2023 and 2024, but then a year ago they changed course drastically. They paused their datacenter construction significantly and slowed down their commitments to OpenAI. We called this out a year ago to datacenter model clients and later wrote a newsletter piece abou
SemiAnalysis
Much has been written about xAI’s Colossus 1. The Memphis build belongs in the history books: the largest AI training cluster, erected from scratch in 122 days. With roughly 200,000 H100/H200s and ~30,000 GB200 NVL72, it remains, today, the largest fully operational, single-coherent cluster (setting
SemiAnalysis
Nvidia announced the Rubin CPX, a solution that is specifically designed to be optimized for the prefill phase, with the single-die Rubin CPX heavily emphasizing compute FLOPS over memory bandwidth. This is a game changer for inference, and its significance is surpassed only by the March 2024 announ
SemiAnalysis
Compute is the lifeblood of AI. He who controls the spice controls the universe the compute will control the production of tokens and reap the benefits of AI. Without compute you do not have a seat at the table. The United States technology community is all in on compute and AI as the next platform
SemiAnalysis
Two-and-a-half years ago, we flagged a looming “cloud crisis” at AWS. Today, the evidence has mounted. AWS is the crown jewel of the Amazon empire, generating ~60% of group profits, and dominating the lucrative Cloud Computing market. But it struggles to translate this strength into the new GPU/XPU
SemiAnalysis
Frontier model training has pushed GPUs and AI systems to their absolute limits, making cost, efficiency, power, performance per TCO, and reliability central to the discussion on effective training. The Hopper vs Blackwell comparisons are not as simple as Nvidia would have you believe.
In this repo
SemiAnalysis
To many power users (Pro and Plus), GPT5 was a disappointing release. But with closer inspection, the real release is focused on the vast majority of ChatGPT’s users, which is the 700m+ free userbase that is growing rapidly. Power users should be disappointed; this release wasn’t for them. The real
SemiAnalysis
The first portion of this report will explain HBM, the manufacturing process, dynamics between vendors, KVCache offload, disaggregated prefill decode, and wide / high-rank EP. The rest of the report will dive deeply into the future of HBM. We will cover the revolutionary change coming to HBM4 with c
SemiAnalysis
Robots have powered manufacturing for decades, yet they stayed single-purpose and thrived only in perfect settings. Previous attempts at intelligent machines overpromised and underdelivered. But they were too early. Today, modern AI paradigms convert most robot roadblocks into data problems and push
SemiAnalysis
Long time readers will recall that SemiAnalysis covers more than just datacenters and AMD. Today we’re back to semiconductors with a tech-focused roundup of the best from this year’s VLSI conference, the premiere design and integration. That includes the latest in chips manufacturing: fab digital tw
SemiAnalysis
Meta’s shocking purchase of 49% of Scale AI at a ~$30B valuation shows that money is of no concern for the $100B annual cashflow ad machine. Despite seemingly unlimited resources, Meta has been falling behind foundation labs in model performance.
The real wake-up call came when Meta lost its lead i