Kumar🇺🇸
42K posts
Kumar🇺🇸
@datarade
dog walker and some other stuff. https://t.co/Y95xEJ8rtj Help folks with growth https://t.co/89nyUiSKYt
Katılım Şubat 2012
7.4K Takip Edilen30.9K Takipçiler
Kumar🇺🇸 retweetledi
Kumar🇺🇸 retweetledi
One kilogram of crystal pulls up to 2.8 liters of drinking water out of dry desert air every day, powered by nothing but sunlight. It's called MOF-801, and the chemistry behind it shared the 2025 Nobel Prize. I build chemical plants. MOF-801 is zirconium nodes linked by fumarate molecules, with pores sized to catch single water molecules. Omar Yaghi's company Atoco is building container-scale units for 1,000 liters of clean water a day, off-grid. Prototypes have already run in the California desert. In Dune they were called wind traps. They remained science fiction for 60 years. The bottleneck has moved from chemistry to manufacturing. Growing these frameworks by the ton, pressing the powder into forms that survive years of use, and building the systems around them cheaply enough to matter.
English
I don’t think ya’ll are ready for Garry’s divorce guy arc.
Who will get custody of Kane?
English
@eastdakota @FabianHedin @Lovable You seem extremely excited and like you got a tan.
NYC sun hitting ya hard.
English
We’re so cooked on the materials supply chain.
Sean@SomethingOnSnow
@datarade Their president has a chemical engineering degree. Probably a lesson there.
English
GIF
English
Kumar🇺🇸 retweetledi
If you already didn’t know, most VCs are genocidal.
Harry Stebbings@HarryStebbings
First term sheet signed for an Israeli company where 20VC is lead. 15% ownership. Two of Israel’s greatest operators. Mega market. True pre-seed. LFG.
English
Kumar🇺🇸 retweetledi
Everybody quote tweeted this with spiritual replies.
😳
Kumar🇺🇸@datarade
Chevron—a $351B giant with every political, technical, economic, and social moat imaginable—decided California isn’t worth the fight. Why do you believe El Segundo is a smart bet for a cost-sensitive industrial manufacturing startup?
English
The feeling is mutual.
My question is completely mathematical and applies less to what you’re doing and more to the capex/opex metals/concrete/equipment intensive asset businesses.
You won’t be spending hundreds of millions or billions on equipment short term and you can operate asset light and still get to 300 people etc…..
There’s an entire class of industrial engineering operations research problems m called facility location planning problems. Think 10M-100M+ tonnes/yr of volumetric product output.
English
You know I admire you and think you’re smart
I know building in California, LA no less, is fraught with potentially business killing headaches that Houston wouldn’t have
I hope my unwavering loyalty to El Segundo is never regarded as grifting or juvenile idealism
And I know I will have to do much outside of California to achieve Rainmaker’s ends, I care about winning more than place
But generations have retreated from the fight, and there is no Galt’s Gultch that will remain unmarred. I’d like to try to turn things around here so the next generation doesn’t need to retreat from Houston
English
Kumar🇺🇸 retweetledi
Two isotopes of the same element, differing by three neutrons out of 238, spin apart inside a rotor moving faster than the speed of sound. That tiny mass difference is the entire barrier between reactor fuel and a weapon. 440 reactors in 31 countries run on enriched uranium: 9% of global electricity, 18% in the United States alone. One machine barely moves the needle: a single centrifuge produces only a handful of separative work units a year. Cascades run hundreds to thousands of them in parallel just to make a dent. Each rotor spins without stopping for a decade. One vibration at the wrong frequency tears it apart. Gaseous diffusion needed a thousand-plus stages stacked across buildings the size of small towns to do this job. A modern centrifuge cascade needs as few as ten stages in series, at a fiftieth of the energy. That collapse in scale is why the technology has spread to nations that could never have built a diffusion plant. Every nuclear weapons crisis of the last half century traces back to this machine. Whoever controls the cascade controls who gets the bomb.
English
@datarade If you would please consult the graph
x.com/zanemountcastl…
Zane Mountcastle@zanemountcastle
the equivalent of YC's first batch photo for hard tech
English
I once briefly met chess world champion @Kasparov63.
Iirc I challenged him to a blindfold chess game and he was like “no, kid” (fair).
I wanted to continue the conversation so I changed topics away from chess where he is 1 trillion times more knowledgeable to politics where the gap was more manageable.
He said something interesting:
“Putin creates conflicts in the Middle East to increase the price of oil, on which Russian government revenues are highly dependent.”
Lot of wisdom there!
English
The Groq acquisition was likely the penalty fee to Nvidia for laundering compute.
My conspiracy theory - co-opted from a friend - and seems credible is that the administration walked Jensen through the White House and told him he needs to buy Groq and administration insiders own a ton of the stock otherwise they'd fuck him up over selling Nvidia GPUs to China illegally through SMCI.
Groq’s projected 2025 revenue: ~$500M
Their LPU (language processing unit) chip taped out in 2020 — already outdated.
Running Llama 70B needs 100's of chips and $M s in hardware.
I've spoken with more than a few chip engineers - I haven't met a single one that thinks that Groq's unrevealed LPU2 can contend with Nvidia silicon teams.
Groq has attrition btw.
Also, massive total cost of engineering.....
The only other plausible weird reason for Nvidia to buy Groq is that Nvidia thinks they'll stockout on their silicon and they'll need to sell something..... at some point.
English
@_LittleFire1 Outdated viewpoint. They're mogging us on novel medical IP.
English
@datarade They sure stole a lot along the way. The Chinese are only good at taking already made things and making them better. Hardly innovators.
English
In <10 years, the Chinese beat Germany, Japan, and the USA automotive industries which were built over 75 years.
They have probably already beaten America at biotech, robotics, and AI.
They're going to be > 50% of novel medical technologies.
They went from < 10% to ~50% in < 20 years on novel medical IP.
English