深潮_TechFlow

Which company has the strongest AI computing power? Take a look at the data center ranking.
A research institute @EpochAIResearch has released a global AI data center TOP60 ranking as of July 24.
For fair calculation, regardless of what chips are used, all compute has been converted into equivalent NVIDIA H100 compute.
The results show that, among today’s AI data centers, the top three by compute power are all in the United States:
Elon Musk’s Colossus 2 in Memphis, Texas ranks first, with the equivalent of 1110k H100s. It cost $35.8 billion to build, far ahead of the rest.
Microsoft’s Atlanta

Which company has the strongest AI computing power? Take a look at the data center ranking list.
Research institution @EpochAIResearchI released a global AI data center TOP60 ranking as of July 24.
To ensure fair comparison, regardless of what chips are used, everything was converted into equivalent computing power of NVIDIA H100.
The results show that currently, the top three AI data centers by computing power are all in the United States:
Elon Musk’s Colossus 2 in Memphis, Texas ranks first, with the equivalent of 1.11 million H100s, built at a cost of $35.8 billion, far ahead of the rest.
M

Let’s take a look at the asset allocation shared by crypto OG Laomao @Imlaomao at the ongoing Hong Kong Money Frontier 2026 conference.
Laomao entered the industry in 2013 and co-founded Coin Capital (INBlockchain) with Li Xiaolai. He invested in a batch of early projects such as EOS, Steem, and Sia.
There aren’t many people in this space who made it through the 2014, 2018, and 2022 bear markets and are still alive—and doing pretty well.
He believes the current reasonable asset allocation is:
40% BTC, core assets
30% AI/Tech, productivity assets
20% cash, liquidity reserves
10% other diversifi

Let’s take a look at the asset allocation shared by crypto OG “Old Cat” at the ongoing Hong Kong Money Frontier 2026 conference.
Old Cat entered the industry in 2013 and co-founded Coin Capital (INBlockchain) with Li Xiaolai. He invested in early projects such as EOS, Steem, and Sia.
People who have made it through the 2014, 2018, and 2022 bear markets and are still alive—and doing well—aren’t that many in this space.
He believes the current reasonable asset allocation is:
40% BTC, core assets
30% AI/Tech, productivity assets
20% cash, liquidity reserves
10% other diversified allocations
You c

If you’re trying to lose weight, you might as well start by stockpiling BTC by using bitcoin:native 😂.
On Reddit, a guy shared his experiment:
Whenever he wants to buy chips, candy, or snacks, he looks at the price, then opens the buy-crypto App and buys the same amount as BTC. A bag of $3 chips becomes $3 worth of bitcoin.
His logic is: this money was always going to be spent on weight-gaining cravings. Instead of swapping it for a few minutes of dopamine and a bunch of calories, why not swap it for a long-term asset?
The best outcome is that you get slimmer and the coin goes up. The

It’s hindsight time—let’s look at 2 sets of data:
1. Who priced Changxin Technology most accurately?
2. Is Changxin’s pricing actually expensive?
Regarding the first question, the neutral range disclosed earlier by brokers was 30 to 45 yuan. In the investment community, some unofficial opinions believed the strongest level that wouldn’t break was a 400% move (43.3 yuan). Both sets are below the actual opening at 49.5 yuan.
The Hyperliquid on-chain quote on July 16 was $7.37, which converts to 49.9 yuan at the exchange rate. 11 days later, the A-share opening was at 49.5 yuan, a difference of 0

Hundreds of users treated ChatGPT as a poison and biological weapons consultation hotline, and some people actually got step-by-step instructions.
The Wall Street Journal disclosed that since the summer of 2025, there have been hundreds of such queries. After bio-weapons and counterterrorism experts reviewed some of the records, they concluded that some answers were accurate and specific enough that an ordinary high school student could follow them.
One user even mentioned in the conversation that they wanted to harm their parents, yet ChatGPT continued to provide answers.
Later, OpenAI banned

Oracle cuts 21k jobs due to losing its AI strategy battle. This is its largest round of layoffs in recent times.
Wait, didn’t the AI strategy fail🤔
Doesn’t that mean they didn’t use AI well, so why are they still laying off🤡
But Oracle’s cloud and AI business is actually surging: cloud revenue +39%, AI orders in hand with RPO reaching $638 billion, up 363% year over year, and it also signed a $300 billion+ deal with OpenAI.
But building data centers has crushed free cash flow to -$23.7 billion, and debt has broken past $120 billion, so they’re borrowing money while cutting staff.
If you use

The chairman of South Korea’s second-largest chaebol, SK Group, Choi Tae-won, will today recalculate a divorce settlement worth nearly $1 billion.
In 2024, the appellate court in the second instance ordered him to pay his ex-wife, Ryu So-young, a property division settlement of 1.3808 trillion won, plus an additional 2 billion won in damages.
The Supreme Court later upheld the 2 billion won in damages but required the property division amount to be recalculated.
The issue lies in a sum of 30 billion won. The second-instance court held that the money came from Ryu So-young’s father, former Pres

At the opening ceremony of the International Congress of Mathematicians held today in Philadelphia, the Fields Prizes were officially announced, and two Chinese mathematicians, Deng Yu and Wang Hong, won the awards at the same time.
The Fields Prize is awarded once every four years and is known as the “Nobel Prize of the mathematics world.” This is the first time Chinese mathematicians have received the award, and there are two in one go. The two are classmates from Peking University, Class of 2007.
However, this award outcome was not a secret two weeks ago.
Someone found that on this conferen

The recent “killing spree” in tech stocks in A-shares and US stocks has brought many people’s suffering into the open—for example, Feng Ge, who rebounds as soon as he clears his position, and then loses money as soon as he buys.
But actually, suffering is an age-old business with internally consistent logic, and it’s time for the editor to tell a story:
In the late 19th century, there was a popular tourism project in New York called slumming (a direct translation: going around poor neighborhoods).
The process went like this: for high-net-worth people with money, they would pay to hire a guide

For fans who’ve recently been beaten up badly by the market, here’s a little story to perk you up:
In Georgia in 2022, a herding dog named Casper encountered an attack on the flock by 11 coyotes.
Casper fought to the death and refused to retreat—after a bloody battle, it killed 8 of them, then jumped over a fence and chased the remaining wolves away.
Most strikingly, after the fight it disappeared for two days. Its owner thought it was dead and posted a missing-dog notice. But two days later it came back on its own—its tail was broken, and the skin on its back had been completely torn off.
Eve

The world is a huge cycle of trading tech stocks.
You’re placing bets on Samsung Hynix, and the guys in Korea are placing bets on China A-share Cambricon.
According to the latest transaction data from SEIBro, under the Korean Securities Depository and Settlement Institute, last week Korean investors went on a buying spree of China A-share technology stocks:
AI computing power leader Cambricon ranked first on this list of Korean-funded A-share stock purchases, with a net buy of $2.86M for the week. Meanwhile, a host of domestic semiconductor core enterprises—such as SMIC, Renkon Technology, Hua