Crypto Miners Are Pivoting to AI Training as the Bitcoin Gold Rush Dries Up

The rush to train massive generative AI models has left companies scrabbling for chips, datacenter space, and reliable access to large amounts of cheap power

Crypto miners are facing a new challenge as the bitcoin gold rush dries up and mining earnings decrease. However, a silver lining has emerged with the increasing demand for artificial intelligence (AI) training, which could provide a new use for their facilities.

Crypto mining involves setting up massive datacenters with specialized computer chips to solve mathematical puzzles that underpin the security of various cryptocurrencies. In return for solving these puzzles, miners are rewarded with some of the cryptocurrency being mined.

Most miners generate the bulk of their earnings from bitcoin. However, earlier this year, an event known as "the halving" significantly reduced mining earnings. As part of the bitcoin protocol, the mining reward is halved every four years to increase the scarcity of the coin. Usually, this triggers a rise in the price of bitcoin, which offsets the halving. However, this time around, the price of bitcoin did not increase as expected, impacting miners' profitability.

Fortunately for these miners, another industry with a voracious appetite for computing arrived just in time. The rush to train massive generative AI models has left companies searching for chips, datacenter space, and access to large amounts of cheap power, all of which many miners already have in abundance.

“It [normally] takes 3-5 years to build an HPC-grade data center from scratch,” JPMorgan analysts wrote in a recent note, according to the Financial Times. “This scramble for power puts a premium on companies with access to cheap power today.”

While crypto mining and training AI are not precisely the same, they share crucial similarities. Both activities require huge datacenters specialized to carry out one particular job and both consume large amounts of power. However, because miners have been in the game for a longer period of time and most AI companies only began attempting to train truly massive models with the launch of ChatGPT less than two years ago, the miners have a head start.

Over the years, they have scoured the country for locations with abundant cheap power and ample space to build large datacenters. More significantly, they have already completed the time-consuming process of obtaining approvals, negotiating power licenses, and getting the facilities up and running.

The rapid demand for AI training is putting a strain on grids in some areas, leading many jurisdictions in North America to have long waitlists for new datacenters, according to Time. Already, about 83 percent of datacenter capacity currently under construction is being leased in advance, says Bloomberg.

This scenario presents a new opportunity for crypto miners, as the biggest bottleneck for many AI companies is finding the hardware to train their models. “You’ve seen a number of crypto miners that were sort of struggling that have actually made a full pivot away,” Kent Draper, chief commercial officer of crypto miner IREN, told Time.

Converting a bitcoin mine into an AI training cluster is not a direct swap. Typically, AI training is done on GPUs, while bitcoin mining uses specialized mining chips from Bitmain. In many cases, however, it's not the chips that AI companies are primarily interested in, but the infrastructure and power access that the mine already has in place.

In June, crypto miner Core Scientific announced that it will host 270 megawatts of GPUs for the AI infrastructure startup CoreWeave. “We view the opportunity in AI today to be one where we can convert existing infrastructure we own to host clients who are looking to install very large arrays of GPUs for their clients that are ultimately AI clients,” Core Scientific CEO Adam Sullivan told Bloomberg.

Some miners are also operating GPUs themselves. German miner Northern Data had previously purchased $800 million of Nvidia GPUs to mine the Ethereum cryptocurrency, but a major software update to the coin’s blockchain in 2022 did away with mining and left those chips sitting idle. Now, the company has repurposed them into a 20,000-GPU training cluster, which is among the largest in Europe, according to Bloomberg.

Other miners, such as Hut 8 and IREN, are making significant investments in new chips to more積極的に pursue the AI boom. In some cases, AI training is being carried out alongside crypto mining. “We view them as mutually complementary,” IREN’s Draper told Time. “Bitcoin is instant revenue but somewhat more volatile. AI is customer-dependent—once you have customers, it’s contracted and more stable.”

這種新趨勢也可能帶來一些適度的環境效益。人工智慧訓練和比特幣挖礦都以其巨大的功耗而聞名，這引起了一些人的擔憂。如果對人工智慧不斷增長的需求只是取代現有的採礦基礎設施，而不是需要新的耗電的資料中心，那麼它可能有助於減少該行業日益增長的碳影響。

然而，對於礦工來說，追尋最新的淘金熱可能是一種冒險的策略，特別是考慮到人們越來越擔心人工智慧產業正處於可能很快破裂的泡沫之中。如果發生這種情況，礦工們已經開始開採的豐富的新礦層可能很快就會乾涸。

The above is the detailed content of Crypto Miners Are Pivoting to AI Training as the Bitcoin Gold Rush Dries Up. For more information, please follow other related articles on the PHP Chinese website!