Market Research on market size and growth

4 Likes

Filecoin currently has about 2 EiB worth of storage available (source).

What would a Safe Network with 2 EiB look like?

Let’s say nodes are 50 GiB on average (number pulled from nowhere).

That would mean we’d need 42,949,672 nodes to reach 2 EiB.

Let’s say there’s 150 nodes per section, that gives 286,331 sections.

If the network is perfectly balanced, that means a prefix length of 18 bits for each section. 18 hops maximum when fetching and storing.


New nodes joining the network would need to download 50 GiB of data. Let’s say they’re all on connections with 100 Mbps download speed and 50 Mbps upload speed (taken from speedtest.net/global-index fixed broadband, which is up:down 92:49).

It takes 1h11m to download 50 GiB at 100 Mbps.

Let’s say the first node joins, then two more join from it, then four more from those, each time doubling the number of nodes that are joining. Probably a bit too ideal but let’s see where it goes. This means each node splits their 50 Mbps upload into 2 x 25 Mbps uploads for the 2 new nodes.

It takes 4h46m to transfer 50 GiB at 25 Mbps.

We end up being net positive after 17 doublings of nodes (we’d have 43,046,721 nodes and need only 42,949,672). If each node takes 4h46m to join, that’s 81h total to reach 2 EiB storage (~4 days).

In reality it won’t be at all so streamlined, and 2 EiB of client data is very different to having 2 EiB of spare space available.

But I was surprised how quickly 2 EiB can potentially accumulate from the exponential growth of 50 GiB nodes under ideal circumstances.

13 Likes

Id guess in all likelyhood nodes will be in the TB, that isn’t great, less nodes needed , but still capped by bandwidth, so would no doubt take much longer to reach 2EiB if you leave the up and down speeds as you applied them.

1 Like

Imo, a 500GB to 1TB average node size is more realistic. Even 4TB has a low barrier for entry now. About the same as 1month of bandwidth from a home internet connection.

6 Likes

Maybe someone could calculate an average based on the free hard drive space numbers: Steam Hardware & Software Survey

Then we can assume a certain % will be allocated for safe network.

It’s at least based on some real life data this way. :slight_smile:

2 Likes

I put my calculations in the google spreadsheet Safe Network growth.

It’s interesting because all the ability to grow comes due to parallelism, so almost all the growth happens in the last two or three iterations of network growth.

The key to getting a large network is always more nodes, not bigger nodes.

The numbers for this are pretty eye-opening (from the spreadsheet):

Node Size (GiB) Iterations to 2 EiB Time to 2 EiB (h)
1 21 2
10 19 18.1
50 17 81.1
100 17 162.3
500 15 715.8
1000 15 1431.7
2000 14 2672.4
5000 13 6203.8
10000 13 12407.7

Onboarding large nodes takes longer so it also takes longer to get to that juicy late stage exponential growth.

Very small nodes mean the network can get the exponential growth cranked up much faster.

I should put a caveat here that the model assumes constant focused growth, when in reality the network will have other demands, has client bandwidth demands as well as node joining, will have varying node speeds, and lots of other factors that make the growth rate not perfectly exponential.

This has really made me reconsider the value of having a fixed target node size. Rather than have elders target 50% full nodes, they could target 50% nodes above/below X GiB (which is really easy to do since they track the chunks for each node anyhow). The average node size has a huge implication on the overall network growth rate.

And I have also come to appreciate how important it is to have a large network for security. More sections means more opportunities for a diverse range of node operators to join. When the network is small it’s harder to join and we have to queue nodes up, which becomes a risk since the smaller operators are probably less inclined to queue, and can also be drowned out of the queue by bigger operators.

Relatively smaller node sizes helps increase parallelism and reduces bottlenecks. Elders can only track a certain amount of churn, so the longer any one churn activity takes the more it blocks other churn from happening. The longer it takes the current node to join the longer it takes before the next two can join after it (the numbers are specific to the model but the idea applies in general).

This is a nice set of stats! The two most common buckets are:

22.83% of users have between 100-249 GB spare space
23.46% of users have between 250-499 GB spare space

50 GiB would be between 10%-50% of spare space for these users. 50 GiB sounds like a reasonable amount in those circumstances.

14 Likes

I have always considered that the desired node storage size would be related to the size of the network (in nodes). Initially thought 2GB for pre-live networks growing to around 8GB when going live and growing from there. While these were figures plucked from thin air, they were molded from experience. Like 2GB is great for small SBCs with 8GB onboard storage, then the SBC can add a USB memory stick for a couple of vaults of at least 8GB.

Thanks for the figures @mav.

But I would consider that people will run up multiple nodes dividing their storage space between them. This I would say yes to

Very small nodes mean the network can get the exponential growth cranked up much faster.

My thought is that smaller node storage size will mean early on the growth will be faster than binary growth (2, 4, 8, 16, etc). I can see some people running up 2 to 5 nodes on each device they use. Using the 100 to 500GB of free storage for nearly 50% of users

7 Likes

Yes, smaller nodes allow greater granularity in disk consumption but will use more cpu because there’ll be more routing to deal with.

Not exactly sure where the ‘best’ balance is.

But the table above makes me err on the side of smaller rather than larger.

7 Likes

Then again, smaller vaults → smaller earnings. How little is enough to compensate for the nuisance of having to keep your machine always on?

1 Like

If you are going to run 24/7 with 20TB NAS, than the only option is to have many nodes on a same machine. I am not sure how many nodes you can run on some low cost NAS with small RAM and inefficient CPU.

If there will be another option to run on network XYZ with one easy set up and better requirements to join why to bother with Safe Network?

1 Like

It would be interesting to compare the economics of different strategies for different jurisdictions.

I think we could get a baseline figure using cloud services, and then compare different locations and strategies with that. It will be interesting to have figures for what we expect for commercially minded (profit paramount) farmers and regular users who may be happy just to earn storage credits.

1 Like

I think it is not so much about being driven by the profit or earning storage credits, but what is the minimum reward that will direct human behaviour in a way that they will setup a node and keep it online. And what are the vectors against setting up a node and keeping it online?

If I would not be one of the people in this for ideological reasons, I would need maybe something like 5 - 10€ / month at minimum to think it is worthwile. Why 2€ is not enough? Well, it is just annoying to have a machine on all the time. And I bet after a month or three, I would need to spend an hour updating something in there: the node itself, or OS or whatever.

If I could operate a node from my phone, I think I would have a small node without worrying too much if it is always connected or not. I would put mild effort to keep in connected, in order to earn for paying my uploads, but if it drops, I would not worry too much. Just live a normal life with my phone. If reconnecting disturbs this normal life too much, for example if it jams the normal function of my phone, I would not do it. In that case I might just buy SNT. The price of it is expected to go up, and the little I buy now would go a long way, maybe rest of my life. No need to bother with node.

My point is, that even though there is excess disk space available, there is no free work from humans. And in this context I define anything that disturbs the habitual “no thought required” -way of living as work. People are not going to keep doing that for pennies.

Of course, when the price is going up, the pennies earned by keeping a node online, keeps growing too. But still the nuisance for 2€ / month (at any current month), is very little compared to just buying that amount from an exchange. As @happybeing said, this threshold is of course very different in different areas of the World. We may see a lot of elders in areas, where connectivity is good enough, and average income not so high.

4 Likes

I think an underappreciated aspect here is expectations of future appreciation of the token. Bitcoin demonstrated this and the importance of economic incentives to a successful and growing network.

1 Like

Interesting thing I noticed with Bitcoin (and expect to be same with Safe) is there is some psychological thing that tells people mining your own crypto is better that buying it. Even when the cost to mine coins is higher at the end, people like to mine their coins.

4 Likes

It isn’t just psychological. There are real benefits:

  • No need for inconvenience of doing KYC at exchange
  • No risk of your KYC info being stolen from exchange
  • No counterparty risk during the buying process
5 Likes

I guess some do. But I have also heard “it made my computer fans run full blast, so I stopped” -stories. I have never mined myself. I tried to mine Monero once, but I don’t remember why I didn’t go on with it. It might be that I just couldn’t set it up, or maybe it made my machine run hot.

Filecoin has just passed 2 EiB of storage space!

Filecoin launched with about 581 PiB of storage on 2020-10-16. 38 days later on 2020-11-23 it reached 1 EiB by adding another 443 PiB, which works out to about 87 days per EiB.

Today on 2021-01-21 it’s another 58 days after reaching 1 EiB, so that’s 33% faster to add the second EiB.

Overall the rate of storage being added since filecoin launched is (1024 PiB + 443 PiB) / (96.5 days) = 15.2 PiB per day. 648 TiB per hour. 11 TiB per minute. 185 GiB per second. Pretty impressive.

There’s currently 1150 miners.

The top five miners have capacity of 74.2 PiB, 72.6, 51.8, 41.7, 35.0.

Filecoin storage cost is currently $0.071/GB/year, 3.86x cheaper than AWS S3.

Average deal period is 313.37 days

Over the last 2500 deals 58,012 GB has been stored.

The total data stored on filecoin network is 3.3808 PiB, so about 0.165% of available storage is being used.

1 filecoin currently costs $22.23 for a market cap of $995,696,363

7 Likes

I am interested in the following information on how it will change over time:

Speculation VS Usage

Usage:

3 806 442 GB / 96.5 days = 39 444 GB/day
$0.071/GB/year / 365 days = $0,000194521 GB/day
39 444 GB/day * $0,000194521 GB/day = $7,672686324 Usage/day (Is this possible to be true? Math is not my strength) :lol:

Speculation:

$307 921 998 (Trading Volume 24h today) / $7,672686324 Usage/day = 40 132 228x ratio

2 Likes

I didn’t check your source figures.
But 0,000194531 should be a . Not a ,
So result should be 7.67225244

Edit.

So you would certainly need to do the speculation sum again using my corrected figure.
Anyhow I’ve just woke, time for coffee.

1 Like

To fill 74.2PiB storage is like 2years with 10Gbps upload stream .

1 Like