Bitcoin miner and data center operator TeraWulf saw its stock fly by 5% after Google raised its stake in the company from 8% to 14%. Why is Google investing so much?
Summary
- Google has upped their investment stake into TeraWulf by $3.2 billion or equal to 14% of the company.
- TeraWulf’s Nvidia GPU may be of use to Google for expanding its AI and machine learning engine, Gemini.
TeraWulf announced on August 18 that Google has raised its stake in the company in order to support its data center expansion at the company’s data center campus in Lake Mariner, New York through the artificial intelligence platform Fluidstack.
According to the release, Google will be providing a backstop of $1.4 billion in debt financing to support the expansion of the data center campus. In exchange, Google will receive warrants to acquire 32.5 million shares of TeraWulf common stock.
Data from Yahoo finance shows that in the past five days following the announcement, TeraWulf’s stock jumped 5%. It is currently trading at $9.38, having gone up by 4.57% in the past 24 hours. Its trading volume also soared from just around 300,000 to nearly 30 million shares in just two days.
With the recent addition of $1.4 billion in debt financing, Google increases its investment to approximately $3.2 billion after Fluidstack decided to expand its lease as part of its 10-year deal with TeraWulf.
The added investment raises Google’s stake in the company from 8% to 14% of the total TeraWulf equity ownership. This funding will aid Fluidstack, a Google-backed firm, in developing the data center at TeraWulf’s Lake Mariner campus in New York.
Google’s investment into TeraWulf represents a shift in AI computing, considering Google has made strides to develop its AI and Machine Learning Products, particularly Gemini.
TeraWulf’s Bitcoin mining GPUs used for Google’s AI?
Google’s $3.2 billion investment into TeraWulf reflects a growing trend of Bitcoin mining companies using their infrastructure to meet surging AI compute demand. In the case of Bitcoin mining, advanced miners typically use Application-Specific Integrated Circuits or ASICS instead of general-purpose GPUs.
However, there are some GPUs that are used for both crypto mining and AI computing, most of them manufactured by NVIDIA. These include models like the NVIDIA RTX 4090 or RTX 6000 Ada which are popular choices due to their VRAM and >>Tensor Cores, which are crucial for AI, as well as their parallel processing for mining.
Other options include the NVIDIA RTX 3060 at 12GB and older models like the NVIDIA GTX 1080 Ti or >>RTX 2060. In the second quarter of 2024, TeraWulf announced that it was committed to purchasing a 128-GPU cluster from NVIDIA.
While the specific NVIDIA model wasn’t disclosed, the company said that this cluster supports its nascent high-performance computing and AI initiatives. This means that Google’s Gemini will be able to run on TeraWulf’s NVIDIA GPUs, which are usually meant for mining crypto.
Although, cryptocurrency mining hardware becomes obsolete relatively quickly; often within a 1.5 to 3-year timeframe due to the rapid technological advancements. In this case, the technology could still be of use to AI and Machine Learning platforms like Google’s Gemini in the long run.
Source link
