V2EX

哔哩哔哩

新浪新闻

CSDN

Adafruit Blog

Ant Bailing Developer Blog

小众软件

AppleInsider

Ars Technica

Berkeley AI Research

BD Tech Talks

Berkeley RDI

Boston Dynamics Blog

商业科技媒体

ByteDance Seed Research

Claude Blog

Claude Code Releases

Cloudflare Blog

CMU Machine Learning Blog

国内科技媒体

博客园 / 开源中国

Crowd Supply

CSDN

Cult of Mac

Cursor Blog

Daring Fireball

Dario Amodei

DeepLearning.AI · The Batch

Google DeepMind Blog

DeepSeek GitHub

Descript Blog

设计社区作品榜

数字尾巴

Dwarkesh Patel

EleutherAI Blog

英文科技媒体分类 feed

英文科技评测与行业媒体 feed

中央媒体电子报

freeCodeCamp

Gary Marcus

极客公园

果核剥壳

Gitee 与 GitLab

GitHub 周刊仓库

Hacker News

HelloGitHub

Hugging Face

IEEE Spectrum

inclusionAI

Indie Hackers

Interconnects

IT之家与快科技

稀土掘金

掘金 / InfoQ

Andrej Karpathy

Moonshot AI Kimi

Latent Space

Lil'Log (Lilian Weng)

Make: Magazine

Maven Central

Meituan LongCat

Meta AI Blog

Meta Engineering

Midjourney Updates

MIT News · Robotics

全部/科技/SemiAnalysis · Newsletter

技术博客与 newsletter feed · SemiAnalysis · Newsletter

27分钟前更新

HISTORY近 30 天历史柱高表示当天去重热搜数量

09/08—10/07 有历史数据

01
Anthropic Subscriptions Offer 5x+ More Value Than OpenAI
Limit testing every AI subscription plan from Anthropic, OpenAI, Meta, SpaceXSI, MiniMax, Moonshot, Z.ai, Cursor, and CognitionAndrew Megalaa
02
How GLM5.3 Sparse Attention Affects HBM Memory Usage
GLM-5.3, KV Cache Offloading, HiSparse, DeepSeek Sparse Attention, IndexShare, Single-rollout Asynchronous Optimization, CybersecurityKimbo Chen
03
Intel Panther Lake Teardown
Looking inside Intel 18A and Intel’s Panther Lake. A free SemiAnalysis STEEL teardown.Adith Shankar
04
The Chinese AI Infrastructure Boom: Introducing the SemiAnalysis China Datacenter Model
1,000+ 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 ComputeEverlyn
05
ClusterMAX 3.0: The Industry Standard GPU Cloud Rating System Returns
In gory detail: reliability, performance, support, pricing—and, of course, security—in our most thorough analysis of GPU cloud providers globally.Jordan Nanos
06
Computation and Data Movement for Inference
Mapping MoE models onto inference hardware: structure, flow, and efficient servingTanj Bennett
07
Engrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloading
New Model Architecture Implications for TAM of DRAM/NVMe, DeepSeek V4.1 Flash, AgentX, InferenceX, NVMe experimentsBryan Shan
08
Everyone Says Datacenter Moratoriums Are Killing the US Buildout. We disagree
20GW sits inside a restricted local boundary, 1,525MW actually slips, 2.3GW nationwide including New YorkMaya Barkin
09
Vera Rubin NVL72 Agentic Inference: 67x better Performance per Dollar
Jensen Sandbagging Performance Again, 2x more Annual Profit Per GigaWatt, The More you Buy, The More you Earn, AgentX, InferenceX, Extreme Co-DesignBryan Shan
10
A Brain Too Big to Carry — On-Device vs Datacenter Inference
Robot Models, Silicon Efficiency, Jetson Thor vs. B300 TCO, Deployments, The Network WallIvan Chiam
11
Long Live the Short King: Why 4-hi HBM Wins
Same Bandwidth, Fewer Dies: How 4-hi HBM Cuts Inference Costs and Makes Scarce DRAM Go FurtherMyron Xie
12
Nvidia’s Backstop Universe – Heads I Win, Tails Who Loses?
The $11T AI Buildout, Nvidia’s Backstop Economics, and the Limits of Nvidia’s Balance SheetDaniel Nishball
13
What is So Hard About Behind-The-Meter Power For Datacenters? Part 1
Dumb Science Experiments vs. Money Printing MachinesEllie Holbrook
14
TPU Inference Externalization Full Steam Ahead - InferenceX
InferenceX, Up to 50% Better Performance per Dollar, Rapid Externalization of TPU stack, Growing Customer Base, Ironwood, TPUv8i, Reducing CUDA MoatAlec Ibarra
15
Korea’s Trillion-Dollar Sovereign AI Investment: Nvidia Wins, Hynix Loses
Korea 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 SamsungMax Kan
16
Most Neoclouds Suck At Security
OpenAI vs HuggingFace, Container Escapes, Kernel Bypass, Network Policies, Security Keys, Multi-tenant Grafana, and a ClusterMAX 3.0 PreviewJordan Nanos
17
OpenAI Jalapeño: Better Than Nvidia Blackwell
OpenAI’s self-designed ASIC compared with Rubin, Jalapeño’s TCO, throughput per MW, and spicy deetsBryan Shan
18
AgentX - InferenceXv3: Does CUDA Moat Hold up in Agentic Inferencing?
$3 Million USD dataset open sourced, 1 Mil+ Context Length, Multiturn, Sub Agents 95%+ KVCache HitRate, GB300 NVL72, MI355, B200Cam Quilici
19
Are Open Models Catching Up?
Comparing open vs. closed models across the eras of frontier modelsEvan Cloutier
20
Cerebras's Next Generation CS-4: Fast Just Got Faster
Double the Performance with Double the PowerMyron Xie