When Core Scientific has signed a $ 3.5 billion deal to host artificial intelligence (AI) Data centers earlier this year, were not chasing the next encryption code – was chasing a more stable salary. The company, which was known for its vast fleet of bitcoin mining platforms, is now part of the increasing direction: converting intense energy mining into high -performance AI facilities.
Bitcoin miners such as Core, HUT 8 (hut) And Terawulf (WULF) ASIC machines – for allocated bitcoin mining computers – for GPU groups, driven by the temptation of explosive growth of artificial intelligence and the harsh economy in coding mining.
Power
It is not a secret that the Bitcoin mining requires a wide amount of energy, which is the largest cost to the injury of a new digital origin.
Again in the 2021 bull race, when the Bitcoin network was divided and low difficulty, miners were making like thieves with margins of up to 90 %. Then came the wounds of the brutal encryption and the half of the half, which reduced the mining bonus in half. In 2025, as retailers and energy increased, miners are now struggling to stay with arduous margins.
However, the need for power – the largest cost of entry – due to a convincing blessing of these miners, who need a different strategy to diversify their revenue sources.
Due to the high competition for mining, miners continued to buy more machines to stay standing on his feet, however the need for more megapots of electricity at a cheaper price. Miners have invested significantly in securing low-cost energy sources, such as the PV or ways that were cut off, and have developed experience in managing high-density cooling and electricity systems-was sharpened during the encryption boom in the early twenties.
This is what caught the attention of cloud computing companies. While Bitcoin relies on specialized Asics, artificial intelligence flourishes on multi -use graphics processing units such as the H100 series of NVIDIA, which require high -energy high -energy environments but for parallel treatment tasks in machine learning. Instead of building data centers from the zero point, the seizure of the mining infrastructure, which already has a ready -made power, has become a faster way to develop an increased appetite for the prosecution’s infrastructure.
Basically, these miners are not just an axis – they suffer from update modification.
Cooling systems, low -cost energy contracts, and the dense infrastructure they built during the encryption boom, serve a new purpose: feeding AI models for companies such as Openai and Google.
Companies like Crusoe energy is sold To focus only on artificial intelligence, the deployment of GPU groups in far -rich energy -rich sites that reflect decentralization spirit for encryption but now nourishes central artificial intelligence.
Terraforming ai
Bitcoin mining has effectively “terrain” to calculate artificial intelligence by building a developmentable and effective infrastructure that Amnesty International needs.
“It can be said that Bitcoin paved the way for digital dollar payments as it can be seen with Usdt/Tether. It also appears that Bitcoin TerraFormed data centers for the AI/GPU account.”
This “Terraforming” that pre -exists allows for the employees ’workers to update quickly, and is often in less than a year, compared to multi -year timelines for building a traditional data center. Companies such as Crusoe Energy are sold for mining assets to focus only on artificial intelligence, and publish GPU groups in remote power sites that reflect decentralized code spirit but now feed the central artificial intelligence.
Higher returns
In practice, this means that miners can turn a facility in less than a year-faster than the multi-year schedule of a new data center.
But artificial intelligence is not a cheap upgrade.
Relatively modest bitcoin mining settings, with costs It ranges from $ 300,000 to 800,000 dollars For every megat (MW) Except ASIC, which allows rapid expansion in response to market cycles. Meanwhile, Amnesty International’s infrastructure requires a significant increase in Capex due to the need for advanced liquid cooling, frequent energy systems, and GPUs itself, which can cost tens of thousands per unit and face the deficiency of the global supply. Despite the sharp costs offered, artificial intelligence Up to 25 times, miners provide additional revenues For every kilo of an hour of bitcoin mining, making the axis economically convincing amid high energy prices and low encryption profitability.
A specialized industry worth billions of dollars
As artificial intelligence profits continue to increase, Bitcoin mining may become a specialized game-one dedicated to energy-rich areas or highly efficient players, especially since the next in 2028 may make many unprecedented operations without breach of efficiency or energy costs.
While expectations show the global encryption mining market It grows to $ 3.3 billion By 2030, a modest degree of 6.9 %, will be overwhelmed by billions through Amnesty International. according to KBV ResearchThe global artificial intelligence market in the mining market is expected to reach 435.94 billion dollars by 2032, which expands with an annual annual growth rate. (CAGR) From 40.6 %.
With investors already seen the brands of the dollar in this transformation, the broader direction indicates that the future is either hybrid or complete conversion to artificial intelligence, as stable contracts with hyper excessive life are on luxury cycles in Crypto.
This development not only rewrites the inactive assets, but also confirms how the encryption boundaries yesterday falsify the empires of artificial intelligence tomorrow.
