this is exactly why I tell people to NOT get a Mac anything for LLMs useless for concurrency and agents, basically a single chat interface that doesn’t scale up neither with sessions nor with kvcache you want to be ready for the agentic world you Buy a GPU
COMPUTING
-
Why Humanoid Robots Are Extremely Difficult to Build
By
–
Elon Musk Answer On Why Humanoid Robots Are So Hard To Build! https://
youtu.be/kM1Ll-dGdto?si
=BBO5u1TWycSk2_9C
… #humanoidtech #humanoid #robot #Robotics #AI #TechRevolution #TechInnovation #ArtificialInteligence #PhysicalAI @AlbertoEMachado @Eli_Krumova @postoff25 @Khulood_Almani @anand_narang -
Sticking to Predictions: The Importance of the Right Tech Stack
By
–
you guys should talk to @covenantlabsai about encrypting your models so it's truly end-to-end (seems like you currently decrypt the data to use an unencrypted model?)
-
Attention Head as a Differentiable Multiplexer
By
–
An attention head is a differentiable multiplexer.
-
Smart Money Investing in Quantum Computing Leadership
By
–
The Smart Money In Quantum: Where Leaders Are Investing Attention, Talent And Capital
by @pravirmalik @Forbes Learn more: https://
bit.ly/40tdZ6j #QuantumComputing #EmergingTech #Technology #Innovation #Tech -
Green opposition threatens data centre and nuclear infrastructure expansion
By
–
Something tells me those same Greens are going to galvanise to block data centres (and nuclear), as imminent Climate Death will simply be substituted by (P) Doom.
-
Using Claude Code on VPS as a Thin Client Dev Environment
By
–
Got the Neo to try it as a dumb client with only @TermiusHQ installed to SSH and solely Claude Code on VPS No local environment anymore It's a new era (Oh and a pink Neo for gf)
-
GTC Live Keynote: Accelerated Computing and AI Transform Tech Infrastructure
By
–
Thank you to our GTC Live hosts and guests for joining us at the keynote pregame. The show brought together top leaders across tech, infrastructure, and industry to explore how accelerated computing and AI are transforming everything from chip design and data centers to
-
Running 397B MoE Model on M3 Mac with Efficient Weight Streaming
By
–
Dan says he's got Qwen 3.5 397B-A17B – a 209GB on disk MoE model – running on an M3 Mac at ~5.7 tokens per second using only 5.5 GB of active memory (!) by quantizing and then streaming weights from SSD (at ~17GB/s), since MoE models only use a small subset of their weights for
-
XPeng’s IRON Robot Demonstrates Humanoids Ready for Daily Tasks
By
–
XPeng’s IRON #Robot Proves Humanoids Are Ready for Daily Tasks
— Ronald van Loon (@Ronald_vanLoon) 18 mars 2026
by @TheHumanoidHub#Robotics #ArtificialIntelligence #Innovation #Technology pic.twitter.com/PLuJmT3teWXPeng’s IRON #Robot Proves Humanoids Are Ready for Daily Tasks
by @TheHumanoidHub #Robotics #ArtificialIntelligence #Innovation #Technology