2M token context sounds incredible but I wonder how it works in practice. KV cache at that scale is a real engineering problem, and results are quite often disappointing for higher context, especially for inter-connected questions that basically needs some sort of "retrieval"
COMPUTING
-
Quantum Computers Are Coming But Not As Expected
By
–
🚨 Quantum Computers Are Coming… But Not Like You Think Quantum computers have long been "science fiction." They don't just use 1s and 0s — they explore multiple possibilities at once.
Now, for the first time, they're showing real-world power: discovering new medicines, solving puzzles classical computers can't. But here's the catch — they're not ready for daily use. What we see today is just the start… a tiny crack in a door that could change everything.
Are we ready for what's behind it? Source
Marr, B. Quantum computing trends that will shape every industry. Forbes. [Translated from EN to English]→ View original post on X — @deeplearn007, 2026-04-05 12:25 UTC
-
Local 26B LLM Running at 166 Tokens/s on NVIDIA RTX 5090
By
–
Voici c'est quoi 166 token/s sur une NVIDIA RTX 5090 sur Gemma4 , avec le modele 26B – 8 bits (qui est à mon sens largement ok!)
— Defend Intelligence (Anis Ayari) (@DFintelligence) 5 avril 2026
Regardez la vitesse svp. On est en avril 2026 et on a ce niveau de LLM en local à une vitesse incroyable. Je suis vraiment trop heureux. Imaginez fin… pic.twitter.com/uAJBj9VdjAVoici c'est quoi 166 token/s sur une NVIDIA RTX 5090 sur Gemma4 , avec le modele 26B – 8 bits (qui est à mon sens largement ok!) Regardez la vitesse svp. On est en avril 2026 et on a ce niveau de LLM en local à une vitesse incroyable. Je suis vraiment trop heureux. Imaginez fin
-
MacBook Air M5 Sufficient for Coding Agents Work
By
–
I spent 3 hours this morning working with coding agents on MacBook 16" M5 Max in LOW POWER mode! 😱 I noticed 0 difference! This means a MacBook Air M5 is more than enough for this. BTW Apple Silicon is unbeatable! 🤷🏻♂️
→ View original post on X — @clementdelangue, 2026-04-05 08:53 UTC
-
Anthropic’s Claude Mythos Model Faces Efficiency Challenges Before Release
By
–
Quick reminder: As you know, Anthropic accidentally leaked that its next flagship model, Claude Mythos, is so compute-intensive that the company admits it needs to become "much more efficient" before any general release. Sounds like we will see "Spud" before Mythos, although
-
Ray Kurzweil, the Modern Nostradamus of Technology
By
–
Ray Kurzweil is basically the modern Nostradamus of technology. People mocked the exponential curves, the timelines, the confidence. Then reality started catching up. He saw where computing was going long before most people even understood the game. [Translated from EN to English]
→ View original post on X — @ceobillionaire, 2026-04-05 01:07 UTC
-
Digital Computers Were a Mistake: IEEE 754 Floating Point Hack
By
–
Digital computers were a mistake. Yaroslav Bulatov (@yaroslavvb) During last week's visit @stephen_wolfram talked about the roots of ML. Why do we use continuous numbers? He tracked it down to 1943 paper. At the time analogue computers seemed like a feasible option. Using them on digital computers needs IEEE 754 which seems like a giant hack — https://nitter.net/yaroslavvb/status/2040462121734180907#m
-
Energy-Efficient Sensor Networks in IoT Systems
By
–
@antgrasso @KirkDBorne @Ronald_vanLoon @IIoT_World @CRudinschi @agentic_factory This thread touches on energy-efficient sensor networks. Worth a look.
-
Balancing AI Model Accuracy with Minimal Energy Footprint
By
–
High-level monitoring without battery drain addresses a key pain point. Hey, community, how do you balance the model's accuracy with keeping that footprint truly tiny?
-
Jump Desktop and VNC Setup for Mac Studio Remote Access
By
–
Install Jump Desktop and vnc into a Mac Studio in your office. Yes there's 1000 other "better" ways like tmux but this won for me because it's just sooooo convenient. And your clanker will run no matter what.