As artificial intelligence (AI) rapidly changes one sector of the economy after another, there is an urgent demand for compute resources to fuel all of this machine intelligence. In addition to metaverse news, we have been reading more news on AI recently.
Even before traffic reached its current peak, hosting the early ChatGPT demo cost OpenAI over $100,000 per day. It costs more than $5 million to train a model like ChatGPT. AI is more than just text generation; it can be utilized to solve real-world problems across a wide range of sectors. To achieve this, comparable large neural models must be trained on a wide range of data sources, including financial, geographic, consumer, and medical data. The shift from current neural net AI’s limitations to systems with higher levels of artificial general intelligence would almost certainly need significantly more computer capacity.
It seems to reason that a small but rising number of bitcoin miners are now thinking about how they may leverage their own computer infrastructures to accelerate the AI revolution.
Bitcoin mining is still a profitable business. Mining other cryptocurrencies is still profitable, despite the market’s quick changes. For example, ether miners faced a big loss when the Ethereum network transitioned from proof-of-work to proof-of-stake late last year.
As a result of the technical and economic conditions in the cryptocurrency business during the last two years, an increasing number of crypto mining companies are investigating the prospect of employing their facilities for other purposes, such as high-performance computing and, in particular, AI.
High-performance computing (HPC) and AI processing frequently necessitate specialized computing hardware that is not optimal for bitcoin mining. Yet, setting up a mining farm isn’t always the most difficult aspect of purchasing servers. The physical infrastructure, which includes electrical power, cooling, security, and other components, necessitates a large investment. This is true whether RAM-light GPUs are used for Ethereum mining or RAM-heavy GPUs are used for AI model learning.
Hut 8, a mining company, has established a precedent by adapting previously mining-specific computing resources for machine learning and other HPC applications. Hive Blockchain has been stocking its servers with processing cards that “can be used for cloud computing and AI applications, as well as rendering for engineering applications and scientific modeling of fluid dynamics” for some time.
Content Source: cointelegraph.com
Cover Image Source: albawaba.com
IMPORTANT DISCLAIMER: All content provided on this website, any hyperlinked sites, social media accounts and other platforms is for general information only and has been procured from third party sources. We make no warranties of any kind regarding this content. None of the content should be interpreted as financial, legal, or other advice meant to be relied on for any purpose. Any use or reliance on this content is done at your own risk and discretion. It is your responsibility to conduct research, review, analyze, and verify the content before relying on it.
