Growth will continue until it is economically infeasible to add another disk to the net, when it is not likely that the new disk will pay off. Let’s say 1TB cost $22 and the disk should pay off in 12 months (depreciation, wear and tear). That’s 1.66 per month and TB. At chiacalulator we are at 9.91. Asuming fixed chia price, netspace can grow by factor 9.91/1.66 ~= 6x and growth be zero threre. 6x would be 100 EiB, just what Bram predicted.
3 Likes
I was working out my total cost per plotted TB yesterday. This would seem to be the single greatest factor ultimately determining network growth.
With the 100EiB I think he was having in mind hordes of people joining with just their free space and low or zero costs. Nobody will do dedicated farming for just a 12 month ROI. That said, if price goes up so does profitability and the 100EiB could not be discarded. I don’t think 100 EiB could be supported with less than $2K coin price with mostly dedicated farmers like we have nowadays.
This is the best explanation I have seen:
https://www.reddit.com/r/chia/comments/np10sn/netspace_is_consistently_slowing_down_what_are/
2 Likes
yes, It’ll stop at 20 zeta or iota bytes!!!
Actually, people expecting 12 Months ROI exist and they will continue to buy netspace.
At my workplace we monitor ebay GPU prices and their possible per day Ether profits (we want to buy GPUs for a machine learning project and believe that Ether still drives prices up).
The ratio between GPU price and per day profit after operating costs is currently between 300 and 500, ~1 year ROI for Ether right now.
Option 2 sounds a bit strange – from my understanding all valid plots are equal. 
So TLDR:
The netspace graph is derived not from counting plots but from an estimate based on the solutions per hour submitted at the current difficulty level, and the seesaw pattern is the result of this calculation feeding back into the difficulty adjustment algorithm oscillating between over and underestimates.
it will only stop once chia is not longer profitable.
@bertybassett I totally agree, but with say $700 a coin where is that end of profitability?
Just an example, say you are running 1PB of space, that is about 72 drives with 14TB, which will fit into 2 cases 4u with 36 cages each. 72 drives at 6W power plus the processor and other stuff around it costs you to be conservative 20 kWh per day. If you have cheap electricity that will cost you about 2$ per day. The investment cost, repair cost, maintenance cost, and ROI is neglected here for simplicity.
So with 1PB space, network at 18 EiB today, the income at the moment is 0.5 XCH per day, say about 350$ per day at $700 per coin.
If you turn this around, the network can still grow about 175 times until 3150 EiB = 3.1 ZiB for a rig like this to be profitable in terms of electricity (not investment cost)
So from my point of view 20 EiB will not be the end of this game if the exchange rate XCH stays as it is or increases.
Another endpoint is the value of XCH, if that were at 4$ per coin the 1PB installation would be at break even in terms of electricity cost.
the good news is yes its slowing however the bad news is it going to explode again when official plotting comes. AKA the second wave.
@bertybassett absolutely, that coming too.
I look forward to how and where this journey will be going. It is unfortunate that it seems to be impossible to rent computing resources to offer plotting services.
another line of thought, assuming you can plot 50 plots per day, exponential growth of the network until 3.1 ZiB you can make only $784 in 6 months time
if you can make 200 plots per day, you will earn $3120 in 6 months
Hmmm, to make 50 plots per day you need about $4k equipment…
just a background, I used advanced https://chiacalculator.com/ with 10 plots starting, plotting speed 5 and 20 TiB per day, unlimited capacity, initial network 18.4 EiB, 40% weekly growth, unbounded, 77 days of exponential growth, 60 days stabilization, 5 PiB per day growth during stabilization
It is a exciting and very dynamic development that chia is taking, breathtaking
IMO: 100TiBs today will be rewarded with around 10 Chia over next 3 years.
Depending on the price, You will earn that much.
There are many offering plotting services but I have not seen one explain how they are taking care of key security, file integrity, or a good guarantee of delivery for payment. I would also be concerned with download time and bandwidth. Perhaps easier if they mail the drives, which I have seen.
I recommend plotting yourself. In the long run, storage is your largest cost, not the plotter.
I bought a refurbished Optiplex 7020 kept the 512GB OS drive, added a Samsung 2TB (1.81TB) 970 EVO Plus for my plotter and a 4TB Samsund SSD (2TB would have been fine) for a secondary plotter.
I plot 3 parallel K33s every 17.5 hours. I could plot 7 K32s in almost the same time, but choose to make K33s for their more long term value. (I also also shave a little time creating K33s, contrary to comments I have read.)
Total cost of plotter was under $2000. 64TB of storage (8TB drives) with drive arrays gives me room for 256 K33s cost about $2500. I can fill this in months and will then make a decision as to whether to continue to expand storage.
The NVMe should last for over 500 K33s before replacement is necessary so this turns into a small expense. If I turn this machine into a farmer after if produces 512 K33s, and add another 64TB my total cost for 512 K33s will be almost exactly $7000, or about $14 per K33 (that would be under $7 per K32 if I were producing them).
I have seen plotters offering K32s for $1 each, delivered via HTTP, but you still need to buy the storage to put these $1 plots on and have a farmer setup … I would have many other concerns about “my” plots.
I considered looking for plots to buy to fill the space I have bought immediately but the numbers just do not work. If I do not plot myself I destroy my (possible, lol) profit margin.
If you are rich enough to buy mass storage and pay to have someone else fill it I can see it working out but you would have to have deep pockets.
I see the point, but see that kind of ROI more for established coins where you can look back for past profitability and project that into the future. Here we are only at the very beginning of the project, no history and no clue of what is really coming.
Also you have to take into account the cost of opportunity, the profit you can make moving your $$$ elsewhere. Here you have a lot of uncertainty, more risk but I have to say also more profit potential for that reason.
Against this last argument is the fact that Chia is monopolistic now with regards to PoST blockchains, so you have available as much storage as you want, on the contrary you don’t have as much GPUs as you want to mine ETH or whatever. So there is no barrier apart from having the $$$ to become a whale overnight.
EDIT: sorry, I replied to the wrong message
The answer to the question posed is no. Since we just passed 21 and still climbing.
I thought I’d gravedig this thread as netspace drops back to 20Eib.
I basically finished plotting unless storj compressed plots or nossd plots
Is anyone still plotting?
1 Like
I keep trying to tell myself I’m finished!
First I thought I would plot until my SSD died. 10 PiB & was going strong so needed another plan
Aimed to finish at 5000 plots but still had a few spare drives
Now at 7000 plots - slowed down significantly but still buy the odd second hand drive if it is good enough value
I think I might be an addict!
1 Like