Futures
Access hundreds of perpetual contracts
TradFi
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Launchpad
Be early to the next big token project
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
More
The article from 36kr provided the following data in the image.
These temporary miners who divert mining hash power to run AI $BTC are no different from retail investors cutting losses at the bottom to buy gold and silver.
They are all chasing gains and selling off on dips, with no independent judgment. Is the US’s shortage of computing power due to a lack of mining machines? No, it’s an energy issue. Nvidia’s new GPUs are about to double their hash rate again and again and again.
Mining machines are not the bottleneck; the generators in mining farms are what AI giants are eyeing. It’s well known that tokens in the US are expensive; spending 1 kWh on tokens can earn you the equivalent of 10 kWh in profit. But tokens won’t stay expensive forever. Don’t you find it frustrating that Minimax is leading in total token consumption? The competitors across the Pacific are priced at only 1/15 of that, and OpenAI and others will eventually have to adjust their prices.
So, the huge profit of earning 10 kWh from 1 kWh of electricity is just a trap, like the play in DeFi mining. Once suppliers are all in place and have invested heavily upfront to build generators and data centers, they will definitely go crazy with the profits.
The next step might even be running tokens at a loss. You have to mine tokens; doing nothing means equipment depreciation just eats away at your assets, and mining at a loss might still recover some of the initial capital.
Or maybe go back to mining BTC? The availability/network latency requirements for token mining are different from BTC mining, requiring additional investments, such as slicing large models, parallel processing, and pipelines that need ultra-low latency and huge throughput between GPUs across multiple servers, with special cabling. 2C inference also needs a high-speed network environment.
These may sound trivial, but the costs of doing them are not cheap at all. After all, American infrastructure relies on private funding, and if your cluster is in a remote area…
In summary, I estimate that after taking a walk around, the bottom-cutting $BTC will have to mine again, mining cards will be iterated, and old machines will lose their advantage and can no longer mine at the current cost of $BTC …
These temporary miners who switch from mining hash power to running AI $BTC are no different from retail investors cutting losses at the bottom to buy gold and silver.
They are all chasing gains and selling off at dips, with no independent judgment. Is the US’s hash power shortage due to your mining machines? No, it’s an energy issue. Nvidia’s new GPUs are about to double their hash rate again.
Mining machines are not the bottleneck; the generators in mining farms are what AI giants are eyeing. It’s well known that tokens in the US are expensive; spending 1 kWh on tokens can earn you the equivalent of 10 kWh in profit. But tokens won’t stay expensive forever. Don’t you envy Minimax leading the total token consumption? The competitors across the Pacific are pricing at only 1/15 of that, and OpenAI and others will eventually have to adjust their prices.
So, making 10 times the profit from 1 kWh of electricity, just like DeFi mining, is a trap set for you. Once suppliers are all in place and have invested heavily in building generators and data centers, they will definitely go crazy with the profits.
The next step might even be running tokens at a loss. You have to mine somehow—letting aging equipment sit idle just results in depreciation to zero, and mining at a loss might still recover some of the initial capital.
Or maybe go back to mining BTC? Running tokens requires different availability/network latency than mining BTC. There are additional investments, such as slicing large models, parallel processing, and pipelines that need ultra-low latency and huge throughput between GPUs across multiple servers, requiring special cabling. 2C inference also needs a high-speed network environment.
These may sound trivial, but the costs of doing them are not cheap at all. After all, American infrastructure relies on private funding. If your cluster is in a remote area…
In short, I estimate that after taking a walk around, the bottom-fishing $BTC will have to mine again, mining cards will be iterated, and the old machines will lose their advantage and can no longer mine at the current cost of $BTC …