₿Bitcoin Radio

Stories · as read on the Mining station

Hive's Three-Cent Floor

Hive Digital Technologies' COO explains why hashprice, not Bitcoin price or difficulty alone, drives equipment decisions, and why Hive targets 25 J/TH.

Oct 16, 20263 min read · 455 wordsRecorded Jun 24, 2024Luke Rossy

Hive Digital Technologies was making one hundred and fifty thousand dollars a day mining Ethereum when the merge hit in September 2022. Its chief operating officer, Luke Rossy, told that story at Mining Disrupt Miami, recorded June twenty-fourth to twenty-sixth, 2024, on the way to a harder question: what should a Bitcoin miner buy, and when? Bitcoin traded at sixty-one thousand, two hundred seventy-two dollars when he spoke. On October sixteenth, 2026, it was eighty-two thousand, five hundred thirty-one dollars.

Rossy's answer was hashprice, which he defined as how much the network is going to pay you for a specific amount of hash on a given day. He said you cannot just look at Bitcoin price or just difficulty and make decisions. In November 2021, with Bitcoin at sixty-five thousand dollars, it took one-eighth of the hashrate it took at the time of his talk to mine a single Bitcoin. Since 2020, he said, hashprice had tested a floor of about five to six cents per terahash per day again and again, and after the 2024 halving it fell below it.

Older machines unplug and difficulty adjusts, but miners keep plugging in equipment bought months ahead, and each new generation of ASICs can mine at lower prices. His prediction: a new floor at three cents a terahash before the 2028 halving. Hive had already planned for it. The company set a fleet target of twenty-five joules per terahash, which puts its break-even at two point seven cents.

It bought more than thirty thousand ASICs to get there, and it runs a bit over five exahash on one hundred forty megawatts in three countries, with room to draw down to one hundred megawatts. Rossy said it is sometimes smarter to buy a generation or two behind, because what matters is how fast a machine pays for itself, ideally four to six months. Then there is power. An S21 brings in maybe one hundred to one hundred thirty dollars of revenue per megawatt hour.

Next-generation AI data centers, Rossy said, are in the thousands of dollars per megawatt hour. In his words: we can't compete with them for the same power resources. He expects Bitcoin miners to move to more remote sites with less reliable power, and that is part of why Hive put its GPUs to work on AI. His last point was fees.

After the 2028 and 2032 halvings, he expects most miner revenue to come from transaction fees, and fleets that run around the clock today may one day switch on only for short windows when a fee rally lifts hashprice. The full story is on Mining Disrupt at miningdisrupt dot com, and many more like it play around the clock on Bitcoin Radio.

More on the air