DeepSeek Shakes Crypto Sectors, Crypto Miners at Potential Loss

DeepSeek Shakes Crypto Sectors

Global financial markets have dropped following the release of a new AI (artificial intelligence) model from Chinese startup DeepSeek. The meltdown erased trillions in combined market value and raised questions about high spending plans on AI infrastructure.

However, the latest addition from DeepSeek named DeepSeek-R1 is a reasoning-based AI that performs better than U.S. based AI models. But it is at a fraction of their development costs. The new model was developed for an estimated $6 million within the span of two months, in sharp contrast to the billions that US tech giants spend on similar A.I. projects.

This gap created a crisis of confidence among investors who had made large and profitable investments on the growth potential of A.I. start-ups. In days, the value of the U.S. stock market was said to have evaporated by over $2 trillion. The cryptocurrency market had its own fall, dropping roughly $300 billion. These two losses combined signaled that expensive AI infrastructure is warranted with a cheaper, open-source option.

For months, the crypto industry had been following trends in the broader technology market, major financial institutions expressed growing interest in digital assets. When tech stocks fell, cryptocurrencies tumbled too. Bitcoin dropped about 3% at one stage, trading just above $102,000. Ether and solana were also down, along with double-digit losses in tokens associated with Artificial Intelligence-related activities.

On top of that, traders suffered nearly $1 billion in liquidated positions in crypto assets, suggesting that happened by the market’s sudden swing. Several participants pointed that the relation between digital assets and tech companies is tight, leading to crypto sharing in the fallout whenever technology stocks fall.

Crypto miners providing HPC services have been particularly exposed. The companies were going to use their computing power for training AI models.This would generate a whole new income stream for the companies. 

However, DeepSeek-R1 raised questions on whether the AI industry needs a large, traditional data center. If sophisticated models can be developed with minimal hardware and lower costs, demand for outsourced computing services could be less than originally estimated. Consequently, shares of crypto mining firms providing HPC hosting, including Core Scientific and others, reduced by 25% to 30% in a matter of days.

The stock of Core Scientific, one of the largest Bitcoin mining companies in North America, dropped by 29% following the news of DeepSeek-R1. Despite the fall, it has been finding a unique path through its mining and hosting segments. The company’s revenue for the third quarter of 2024 reached $95.4 million, which was about $17.6 million lower than the same period a year earlier. During that quarter, Core Scientific mined fewer Bitcoin, faced a net loss exceeding $450 million. At the same time, operating losses was over $40 million.

Before DeepSeek-R1, many technology firms were showing interest in AI projects. Major processor makers were increasing their valuation, as AI Development would need a more processing and computing power. Even cryptocurrency miners used this opportunity to use their data centers for AI workload.

But many data observers believe this fall as a part of the hype. They feel that more AI models like Deepseek can be developed on low value data centers, but high scale AI models will still need computing power, so the sudden dip of the stock market will soon grow.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top