
X - NVIDIA Inception Dev | web5 | $JUP $WPOND $SHIB #NVIDIAOMNIVERSE 🟪💲https://teapawt.net/t.co/LHPx0WHNBi
Dallas, TX
Joined May 2021
- Tweets2.4K
- Following1.4K
- Followers471
- Likes3K
C4P retweeted
It's seriously wild to see my giant mods getting this level of exposure thanks to Project Reclaimer.
I know the gameplay is imperfect but it makes me really happy that my instinct to make massive-scale maps like this was on point and Halo fans are resonating with the vision.
C4P retweeted
They be like, “Yo, I want that old Waka. Go back to that old Waka.”
What you mean, go back to that old Waka? The Waka that used to pull up 100 deep and stomp niggas out? The Waka that used to shoot at people? The Waka that would slap the fuck out of somebody anytime they said something crazy?
You want the sound of the old Waka. Think about the life you asking me to go back to.
Ain’t shit changed about who I am. I just evolved.
Then you hear, “Man, bro be talking all day.” A lot of that comes from young adults who only know what they see online. Your big brothers, your uncles, the older people know me. They know what I lived through. Some of what you hear happened. Some of it’s exaggerated. But I don’t need to keep proving my past by repeating it.
For you young niggas running around misguided, misunderstood, lost, feeling like nobody got you, I want you to see where growth can take you. I evolved because I wanted to be a better leader, a better example, a pathway for y’all. For my family, my friends, my kids, and ultimately for me.
One day I had to ask myself, “Waka, why you keep chasing money when you don’t even understand it? What is money? How does it work? Where does it come from? Why was it designed? Is spending it all you know how to do?”
That’s where financial literacy started for me. With questions. I knew how to make money. I had to learn how to understand it.
A lot of the tools I’ve built for myself came from living, making mistakes, asking questions, and being willing to learn. Now I want to pass them along to the people I love.
I know I’ve been writing a lot lately. But if you follow me and take the time to read, something I share might help you find your way through the chaos you’re dealing with mentally, spiritually, or financially.
It takes patience. It takes discipline. It takes faith.
That’s the Waka you’re hearing now. A man who lived enough to have something to say, and cares enough to share it.
C4P retweeted
#Pondwater 💧 and #Splash 💦, #Bubbles 🫧 are simply mesmerizing and so rewarding, Drink up 🥤
Geoff.ai
Pond0x.com
C4P retweeted
Got inspired after seeing someone create CSS Surf in browser so I did the same with Bunny Hop as I used to spend hours playing on those servers in CSS/gmod.
Play here 🌐 bunnyhop.lol/
C4P retweeted
Huge thanks @chroma_visions for pick up an edition from RGB Playground Collection, I appreciate it a lot!🫂
Attunement
C4P retweeted
Nvidia has now a big problem.
China built an AI that writes CUDA better than humans experts.
Writing high-performance CUDA kernels for GPUs has always been a black art. It requires a brutal, microscopic understanding of hardware architecture, memory hierarchies, and parallel processing.
Even the best LLMs have traditionally choked on it, completely outperformed by old-school compilers like torch.compile.
ByteDance and Tsinghua University dropped a paper introducing “CUDA Agent.”
It is a large-scale agentic reinforcement learning system that transforms LLMs from passive code-completers into active, autonomous hardware optimizers.
Instead of guessing code in one shot, the agent runs in a continuous, multi-turn loop. It writes code, compiles it, profiles the hardware bottlenecks, catches its own bugs, and rewrites it.
The results completely shatter previous benchmarks.
It didn't just match static compilers. It left them in the dust.
• 2.11× faster than torch.compile • 2.60× faster than standard eager mode
• Achieved a massive 98.8% pass rate on complex kernel generation tasks
It discovered advanced operator-fusion techniques that static compiler heuristics can't even see.
For years, we’ve assumed that while AI can write web apps and python scripts, deep hardware-level optimization would require human engineering for a long time.
That assumption just expired.
The models aren't just writing software anymore.
They are optimizing the silicon it runs on.
We're working on a new game about digging a hole, now with co op for 1 to 4 players, where you have to uncover the secret hidden beneath the beach! 👀
It's called "A Game About Digging A Hole™ Together"
Like + RT = Support indie devs 🔥
A video world model can make a convincing clip and still get the physics wrong.
Our researchers just released Physis-Lang, an open self-evolving framework that adds physics reasoning to video captions. The captions explain why and how a scene unfolds. We use them to fine-tune world models and add that reasoning to prompts when generating video.
Adding physics reasoning to the prompt alone improved NVIDIA Cosmos 3’s PhyGenBench score by 5.62 points without retraining.
Paper and project: physis-intelligence.github.i…




























