Ethereum’s Merge was the joining of the original execution layer of Ethereum, the Mainnet that has existed since genesis, with its new proof-of-stake consensus layer, the Beacon Chain.
This transition eliminated the need for energy-intensive mining and instead enabled the network to be secured using staked ETH.
While that update was a truly exciting step in realizing the Ethereum vision, it left miners out of work with many wondering if their equipment would be useless.
Miners have been increasingly pivoting to AI and Vancouver-headquartered HIVE Blockchain Technologies (HIVE) is one of them. With one of the largest fleets among cryptocurrency miners in the world, HIVE is exploring AI and other innovative ventures to utilize its 38,000 Nvidia data center-grade GPUs.
Fintech.ca sat down with HIVE’s CEO Aydin Kilic to discuss how the company is exploring AI and dive deeper into the future of AI in the cryptocurrency space.
Can you tell me a little bit about the background of HIVE?
AK: HIVE is the world’s first publicly traded crypto miner, traded on the TSX since 2017. We are a leader in blockchain technology with data centres in Canada, Sweden, and Iceland. Initially focused on Ethereum mining using our fleet of GPUs (graphic processing units), we have now expanded into two segments: Bitcoin mining and Artificial Intelligence through GPU compute for AI application and private enterprise cloud services.
Our Bitcoin mining capacity is 3.3 Exahash per second, powered by over 100 MW of clean renewable energy. With this, we can generate approximately 8.8 Bitcoin per day; although this fluctuates on a bi-weekly basis, we consistently surpass the industry average.
Our strategic focus to drive advancements in Artificial Intelligence is fostered by high-performance SuperMicro servers equipped with up to 10 NVIDIA A40 GPUs each and is experiencing rapid growth. Additionally, we note HIVE has 38,000 Nvidia GPUs in its fleet, which are also mining Altcoins, which are converted into Bitcoin. As we acquire more SuperMicro servers, we can apply more GPUs to AI workloads.
Why are NVIDIA data-center-grade graphic GPUs important in crypto mining?
AK: Until September 2022, Ethereum, a decentralized blockchain, was mined primarily using GPUs. HIVE was a leading Ethereum miner with a fleet of more than 38,000 high-end NVIDIA GPUs at our disposal.
GPUs can still be used to mine cryptocurrencies today, but not at the same scale when mining Ethereum. As a result, HIVE has expanded its strategic endeavors to utilize its fleet of GPUs to build HIVE Cloud, an enterprise-class cloud service offering robust servers for demanding artificial intelligence (AI), machine learning, and visual applications.
Can you tell me a bit about the Ethereum merge? How has this impacted HIVE’s operations?
AK: HIVE was a pioneer in Ethereum. To help diversify the business, simultaneously, we invested our cash flow, when possible, into large Bitcoin mining facilities with an emphasis on sustainability. The Ethereum Merge, where the network transitioned from proof of work (PoW) to proof of stake (PoS), was inevitable. However, HIVE utilized its GPU fleet to take advantage of the opportunity.
As the world becomes more accustomed to artificial intelligence, at HIVE, we see endless potential in providing cloud compute for AI through our GPUs.
At the end of 2022, HIVE announced that it has a significant fleet of NVIDIA data-center GPUs. Given that HIVE is no longer mining Ethereum, how will these GPUS be utilized?
AK: Today, HIVE uses its industrial-grade NVIDIA GPUs to meet a growing demand for GPU compute for AI application and private enterprise cloud services, which is what powers breakthrough AI applications such as ChatGPT and Midjourney. The underlying technologies powering the AI revolution require vast amounts of GPUs to train and run.
At HIVE, we believe that GPU compute for AI application and private enterprise cloud services is going to be a large and lucrative market, and are focused on creating our own HIVE Cloud platform to capitalize on this industry.
Do you think that more crypto miners will start expanding into AI?
AK: We’re already seeing this happen. There is going to be an incredible amount of demand for GPU compute for AI application and private enterprise cloud services compute over the next decade, largely due to a new wave of AI technologies like ChatGPT.
Our distinctive approach allows us to stick out in this emerging industry. Unlike many miners who solely focus on Bitcoin mining, our advantage lies in our extensive experience and expertise in GPUs.
What’s next for HIVE? What’s on the horizon?
AK: HIVE is in the data centre business, running efficient, clean, and affordable data centres in North America and Europe. We use internal models to decide what the best uses of those data centres are, and as of today, we are utilizing a mix of Bitcoin and GPU compute for AI application and private enterprise cloud services.
We’re bullish on both the Bitcoin and GPU compute for AI application and private enterprise cloud services technologies long-term, and we are well-positioned to take advantage of the dynamic, digital transformation we’re seeing play out in the market.
Leave a Reply