Cheap power is rarely where the contractors are. Mining farms sit next to hydro dams and in deserts, in places that can hit minus 40 degrees Celsius in winter and 40 degrees in summer. Building one site there is hard; building several, and running them all, is a different problem.
At Mining Disrupt, held online on July 22–23, 2020, Philip Salter, then head of operations at Genesis Mining, spoke about the scaling problems miners run into as their data centers grow and spread across the world. Genesis Mining, he said, had been doing industrial-scale mining since 2013, from small sites to hundreds of megawatts across several climates. Bitcoin was trading at $9,530 at the time; on October 11, 2026, it was $83,079. He split the problem into three parts: buildings, software and people.
The End of the Urban Retrofit: Why Cheap Power Means Remote Complexity
Salter began by distinguishing between the small-scale miner and the industrial operator. For the former, a retrofitted building with manual cooling adjustments might suffice. But as operations scale to hundreds of megawatts, the quirks of existing structures become liabilities. Renting or buying urban space introduces unnecessary costs and forces engineers to work around non-optimal electrical and cooling systems. Salter noted that at larger scales, miners are forced to move to the source of cheap power, which often means remote locations near hydro dams or in vast deserts. These sites are further from the populated areas where skilled contractors and service providers are readily available, creating a logistical gap that standard construction methods cannot easily bridge.
Standardizing the Farm: The Case for Modularized Data Center Containers
To solve the remote location problem, Salter advocated for a shift toward modularization. Instead of building unique structures for every site, which involves lengthy permitting, custom engineering, and high on-site labor costs, he proposed a standardized approach. The goal is to perfect a single mining module, manufacture it with quality control, and ship it to the remote site for minimal on-site installation. This method reduces the dependency on local skilled labor and allows for faster deployment.
Genesis Digital Assets has implemented this strategy internally. Salter described their specific solution, a modularized unit designed to handle the extreme temperature variations found in remote sites. By recirculating hot air from the miners back into the intake, the system prevents the damage that occurs when extremely cold air hits sensitive electronics. This hardware standardization is not just a convenience; it is a prerequisite for managing the scale and distance of modern mining farms.
the best way to go forward with this is to have modularized solutions
Philip Salter · 3:36
we're calling it our air cooled data center container and it's a 40-foot shipping container
Philip Salter · 5:14
Beyond Monitoring: The Strategic Shift to Enterprise-Grade Mining Management
With the hardware standardized, the focus shifts to software. Salter said many monitoring tools come with vendor lock-in: built by a pool, a firmware maker or a hardware maker, they work only with that pool, firmware or hardware. A large farm is never uniform. It runs different device types and patches, customers with special requirements and different cooling conditions in different parts of the floor, so a miner ends up managing several tools instead of using one.
He also drew a line between monitoring and management. A monitoring system collects data and displays it. A management system also sets machine configuration and automates as much manual work as possible. He added that many tools are built by startups rather than large miners, so they miss what operators actually need, and that the bigger the operation, the more it needs remote access instead of someone walking the floor. Genesis Mining had built its own tool, Genesis Hive, and was rebuilding it under the project name Hexa.
many of them have a vendor lock-in actually uh these are not solutions
Philip Salter · 6:53
we're calling it an enterprise grade mining management solution
Philip Salter · 9:15
The Audit Trap: Why Data Retention Is a Scaling Requirement, Not a Nice-to-Have
A critical part of enterprise-grade management, according to Salter, is data retention. A small miner mining for itself can show a screenshot and say everything is fine. A miner with customers, investors, partners or auditors seeking a certification cannot. Someone will ask why the farm's performance dropped five percent in September 2018, and the miner needs a good answer backed by data.
Salter's advice was to start collecting and keeping that data early, before it is needed, rather than discover the gap when the question arrives. Once the data exists, it can drive reporting: the transactions that came in, pool performance, farm performance and what last week's downtime cost. A finance team can export that instead of calculating it by hand.
no there'll be questions like oh back in 2018 in september there was a five percent drop
Philip Salter · 10:40
The Human Bottleneck: Structuring Centralized Expertise for Distributed Operations
The final piece of the puzzle is human capital. Salter noted that while software can automate many tasks, it cannot replace the experts who design the farms and solve complex, unforeseen problems. The challenge arises when a miner opens a second location. The experts who succeeded at the first site cannot be in two places at once. Hiring a full team for every new site leads to excessive overhead and knowledge fragmentation.
Salter proposed a structural solution: separating the on-site maintenance teams from the central expert core. The on-site teams handle day-to-day tasks, such as replacing fans and checking electrical installations. The central core team, composed of specialists in cooling, IT, software, and networking, acts as a task force that supports all locations remotely and on-site when necessary. This structure allows miners to scale without doubling their expert headcount for every new site, ensuring that high-level knowledge is centralized and applied consistently across the network.
Genesis Mining, he said, is structured the same way: central offices where the experts sit together, and sites around the world staffed for engineering and maintenance. Hardware, software and people were the three places he saw scaling break. The conversation continues at Intelligence Disrupt 2027, March 22 to 24 in Irving, Texas.
the solution for the issue that you know experts are not always available in every single farm
Philip Salter · 13:09