Research requested: How much will SAFE Network save electricity compared to corporate data silos that require intensive cooling?

Just occurred to me that when SAFE Network launches and apps and data start moving from the corporate data silos to just devices here-and-there that are connected to the Internet it will likely save quite a lot on electricity.

I base this estimation on the fact that consumer devices rely much more on passive cooling systems, whereas the data silos need intensive 24/7 cooling and this is why I am requesting that someone / some organization should study this issue and make scenarios and calculations about the amount of electricity and CO2 emissions saved when there is less need for the corporate data silos as services can run on inexpensive and passive cooling consumer-grade devices.

If the research does in fact show that basing services in the SAFE Network instead of the corporate data silos helps reduce CO2 emissions, this would be a great thing for the planet and definitely very useful in marketing the SAFE Network.


If we see the home vault take hold then it is pretty self evident that its true.

  • Home vaults use spare resources and extremely small incremental increase in electricity due to increased disk usage.
  • this saves equipment being purchased reducing manufacturing of all the CPUs, MBs, Racks, Cooling equipment, buildings, etc
  • so this includes CO2 and more savings

Now some argue that vaults will mostly be in data centres since few people will be able to run vaults at home. This presents a different scenario where we do not see so much savings and the cost of storage on SAFE will be correspondingly higher due to the much higher costs of using data centre machines for vaults. (combined costs of cooling, buildings, electricity, machines etc)

1 Like

I think there would be no reason for current data centers not to move to safenetwork if it succeds?

If I would run a vault I would invest in some extra equipment to run it 24/7. As I like Safenetwork I think the scenario where the spare space of worlds hardrives is used is a little bit utopian.
Correct me if I’m wrong.

It’s not wrong, but there is a heck of a lot of nuance here. My servers already exist, doing things. My marginal cost is 0. We can’t look at this in any context but a cost-curve. If enough people doing ordinary things contribute, then the marginal cost will be closer to 0 than to that of new equipment, which is admittedly higher than the standard data center cost.


The current situation is not necessarily the future one. I can remember people saying computers would always be too large and expensive ever to be used in people’s homes.

Servers may become as common as routers in homes of the future.

I know Solid is not Safe, but there are already quite a few individuals running Solid Pod Servers on a Raspberry Pi, and there are already companies springing up with which you can express your registration of interest from anyone wanting a plug-and-play Solid Data Pod’s.

No. Datacentres still have extra overheads, a lot really.

  • Aircon
  • Staff
  • Maintenance costs
  • cleaners
  • Buildings are paid off over time and is included in server costs
  • Site rental costs, or mortgage costs
  • Upgrade (ie replacement for computers) of machines. Approx every 2-4 years. Other equipment (including aircon) is upgraded according to a schedule.
  • and more


  • very small incremental cost
  • other costs are absorbed because they are incurred in the normal use of the computer.

So no. Data centre machines will always have a higher cost. What you have to weight it up against is the earning rate. Now the margin is dropping because home internet speeds are rising constantly because last mile links are faster over time and the backbone of the ISPs is increasing faster (per customer). Except perhaps for areas of the USA

Now for those who already (few compared to the world) have servers in data centres then they win because its spare resources in a data centre for them. Except in some/many cases bandwidth costs will increase too.

Good point.

Netflix and other steaming services have cause ISPs to increase their backbone bandwidth to accommodate current and future demand. They also mean that increasingly people are installing dedicated machines from google TV sticks to android/apple steaming boxes to laptops. Most of which can run a vault (maybe small). This is the start of that progress to more powerful machines that could act as servers.


If anyone can farm and benefit greatly from how easy it is, how much incentive there is, etc., then the whole world will outweigh any data centers. I thought it was actually counterproductive and not economical, based on how the Network will work—as per the whitepapers/blueprints/concepts/ideas that have existed since 2013+—to own huge centers. Even if that’s not true, I’d imagine that how ubiquitous the Network will be will balance the scale immensely, electricity be dammed :wink:. Maybe people might consider they don’t need unbelievable China-sized rooms full of farming equipment, if they realize everyone benefits instantly/insanely when the word gets out to the world like wildfire, people have incentive to save absolutely first and foremost, and everyone makes a fortune compared to, well, the Fortune 500… whoever they even are. (Who are they? Who really cares, I say!) In that case it’d only take 1% of the world’s huge data centers to be even NEEDED to fulfill people’s desires. Maybe the truly decentralized/distributed nature of the system/currency has something to do with all this. Anyway this might be off-topic

1 Like